ÔØ ÅÒÙ×Ö ÔØ
Multi-decadal satellite measurements of global volcanic degassing
S.A. Carn, L. Clarisse, A.J. Prata
PII: S0377-0273(16)00003-2 DOI: doi: 10.1016/j.jvolgeores.2016.01.002 Reference: VOLGEO 5734
To appear in: Journal of Volcanology and Geothermal Research
Received date: 4 August 2015 Accepted date: 4 January 2016
Please cite this article as: Carn, S.A., Clarisse, L., Prata, A.J., Multi-decadal satellite measurements of global volcanic degassing, Journal of Volcanology and Geothermal Research (2016), doi: 10.1016/j.jvolgeores.2016.01.002
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT
Multi-decadal satellite measurements of global volcanic degassing
S.A. Carn1,2*, L. Clarisse3, A.J. Prata4
1Department of Geological and Mining Engineering and Sciences, Michigan Technological
University, Houghton, MI 49931, USA; [email protected]
2Department of Mineral Sciences, National Museum of Natural History, Smithsonian Institution,
Washington, DC 20560, USA
3Spectroscopie de l‘Atmosphère, Service de Chimie Quantique et Photophysique, Université
Libre de Bruxelles, Brussels, Belgium
4Nicarnica Aviation AS, Kjeller, Norway
*Corresponding author
Invited review article for Journal of Volcanology and Geothermal Research
ACCEPTED MANUSCRIPT
1 ACCEPTED MANUSCRIPT
Abstract
Satellite instruments have been providing measurements of global volcanic emissions of sulfur dioxide (SO2) since 1978, based on observations in the ultraviolet (UV), infrared (IR) and microwave spectral bands. We review recent advances in satellite remote sensing of volcanic gases, focusing on increased instrument sensitivity to tropospheric SO2 emissions and techniques to determine volcanic plume altitude. A synthesis of ~36 years of global UV, IR and microwave satellite measurements yields an updated assessment of the volcanic SO2 flux to the upper troposphere and lower stratosphere (UTLS) between 1978 and 2014 (~1-2 Tg/yr). The present availability of multiple UV and IR satellite SO2 products provides increased confidence in calculated SO2 loadings for many eruptions. We examine the temporal and latitudinal distribution of volcanic SO2 emissions and reassess the relationship between eruptive SO2 discharge and eruption magnitude, finding a first-order correlation between SO2 emission and volcanic explosivity index (VEI), but with significant scatter. Based on the observed SO2-VEI relation, we estimate the fraction of eruptive SO2 emissions released by the smallest eruptions
(~0.48 Tg/yr), which is not recorded by satellite observations. A detailed breakdown of the sources of measured SO2 emissions reveals intuitively expected correlations between eruption frequency, SO2 loadingACCEPTED and volcanic degassing MANUSCRIPT style. We discuss new constraints on e-folding times for SO2 removal in volcanic plumes, and highlight recent measurements of volcanic hydrogen chloride (HCl) injections into the UTLS. An analysis of passive volcanic emissions of
SO2 detected in Ozone Monitoring Instrument (OMI) SO2 data since 2004 provides new insight into the location and stability of the dominant sources of volcanic SO2 over the past decade.
Since volcanic SO2 emissions constitute a random, highly variable perturbation to the atmosphere-climate system, continued monitoring of volcanic SO2 emissions from space by
2 ACCEPTED MANUSCRIPT
multiple UV and IR instruments to extend the current multi-decadal record is essential, and near- global, geostationary measurements of SO2 may be available by the end of the current decade.
Highlights:
A new synthesis of ~36 years of satellite measurements of volcanic degassing is
presented.
A new publicly available database of volcanic sulfur dioxide (SO2) emissions measured
by satellite remote sensing is introduced.
Satellite measurements of SO2 and other volcanic gases including hydrogen chloride
(HCl) and hydrogen sulfide (H2S) are reviewed.
An updated assessment of the volcanic SO2 flux to the upper troposphere and lower
stratosphere (UTLS) between 1978 and 2014.
New satellite-based constraints on the strongest sources of passive volcanic degassing.
Keywords: satellite remote sensing; volcanic SO2 emissions; volcanic degassing; volcanic eruptions; volcanic plumes; volcanic clouds
1. Introduction ACCEPTED MANUSCRIPT
Since the first satellite observation of volcanic sulfur dioxide (SO2) emissions, made after the major eruption of El Chichón (Mexico) in 1982 (Krueger, 1983; Krueger et al., 2008), satellites have provided unique and critical observations of global volcanic degassing. Prior to the satellite remote sensing era, which for SO2 measurements began in 1978 with the launch of the first Total
Ozone Mapping Spectrometer (TOMS; Krueger, 1983; Krueger et al., 1995) and TIROS
Operational Vertical Sounder (TOVS) instruments (Prata et al., 2003), the variability of sulfur
3 ACCEPTED MANUSCRIPT
emissions from volcanic eruptions was very poorly understood, aside from what could be gleaned from early glaciochemical studies of ice cores (Hammer, 1977; Hammer et al., 1980).
Satellite measurements provided unique constraints on the vast SO2 clouds discharged by the major eruptions of El Chichón in 1982 and Pinatubo (Philippines) and Cerro Hudson (Chile) in
1991 (Krueger, 1983; Doiron et al., 1991; Bluth et al., 1992; Schneider et al., 1999; Constantine et al., 2000; Guo et al, 2004; Krueger et al., 2008), spawning a paradigm shift in our understanding of the petrological origins of ‗excess sulfur‘ emissions (e.g., Luhr et al., 1984;
Westrich and Gerlach, 1992; Wallace and Gerlach, 1994; Scaillet et al., 2003; Oppenheimer et al., 2011). More recently, improvements in instrument sensitivity have extended the capabilities of satellite measurements to include smaller eruptions, tropospheric SO2 emissions (e.g., passive volcanic degassing) and other minor volatile species present in volcanic emissions such as hydrogen sulfide (H2S), hydrogen chloride (HCl), bromine monoxide (BrO) and chlorine dioxide
(OClO) (e.g., Eisinger and Burrows, 1998; Carn et al., 2008b; Theys et al., 2009; Clarisse et al.,
2011; Carn et al., 2013; Theys et al., 2014).
Oppenheimer et al. (2011) review the significance of volcanic sulfur emissions for volcano monitoring and their wide-ranging impacts on several facets of the Earth system, including atmosphericACCEPTED chemistry and climate, terrestrialMANUSCRIPT and aquatic environments, and human and animal health. In the context of satellite measurements, which are global in scope and provide unique observations of the largest volcanic SO2 clouds, the primary motivation for monitoring volcanic SO2 emissions is their potential impact on atmospheric chemistry and climate (e.g., Robock, 2000). Many of the relevant science questions that we can address using satellite measurements of volcanic degassing have been outlined by the Stratospheric Sulfur and its Role in Climate (SSiRC; http://www.sparc-ssirc.org/) working group, part of the World
4 ACCEPTED MANUSCRIPT
Climate Research Program‘s (WCRP‘s) Stratosphere-Troposphere Processes and their Role in
Climate (SPARC) project. Current research foci include: the stratospheric sulfur burden and its partitioning between sulfur species including SO2, aerosol, H2SO4, and OCS, etc.; trends in stratospheric aerosol optical depth (SAOD) and the major SAOD forcing mechanisms; the nature of ‗background‘ stratospheric aerosol; the role of minor volcanic eruptions in modulating the stratospheric sulfur burden, particularly since the 1991 Pinatubo eruption; trends in emissions of aerosol precursor gases (e.g., SO2); and the time-constant for SO2 to H2SO4 conversion in SO2- rich volcanic plumes. In order to address these questions, we require measurements of aerosol precursor gases (including SO2 and OCS) and analyses of their temporal variability. Monitoring stratospheric injections of sulfur by volcanic eruptions is of prime importance, though quantification of tropospheric volcanic sulfur emissions is also needed since tropospheric sulfur can cross the tropopause via other mechanisms such as deep tropical convection (e.g., Sheng et al., 2015).
Continuous monitoring of volcanic injections of SO2 into the upper troposphere and lower stratosphere (UTLS) is also required to assess the potential impacts of, and need for, geoengineering of Earth‘s climate (specifically, the branch of geoengineering referred to as solar radiation management).ACCEPTED Injection of sulfate aerosol MANUSCRIPT precursors into the stratosphere, mimicking volcanic forcing of climate, has been proposed to mitigate the impacts of anthropogenic global warming (e.g., Crutzen, 2006; Wigley, 2006; Robock et al., 2009). Accurate characterization of the time-averaged volcanic SO2 flux into the stratosphere is crucial in this regard, and analyses of volcanic eruptions of varying magnitude provide unique insights into the behavior of SO2 in the
UTLS after injection. Recently, volcanic influence on the post-Pinatubo stratosphere has been scrutinized in considerable detail in order to establish the cause of the reduced rate of global
5 ACCEPTED MANUSCRIPT
warming observed in global mean surface temperatures over the last decade (dubbed the global warming ‗hiatus‘). A particular focus has been the role of ‗small‘ volcanic eruptions, or those with a volcanic explosivity index (VEI; Newhall and Self, 1982) below 5. Although alternative causes and even the very existence of the global warming hiatus are subject to ongoing debate
(e.g., Schmidt et al., 2014; Steinman et al., 2015; Karl et al., 2015), several studies have concluded that SAOD perturbations by small volcanic eruptions not included in pre-hiatus climate model simulations have likely played a role. But regardless of the dominant forcing factors influencing recent trends in global warming, it remains crucial to monitor volcanic degassing and quantify SO2 and other volcanic gas fluxes into the atmosphere.
In this contribution we review the current state of our knowledge of volcanic degassing
(primarily focused on SO2 emissions) based on multi-decadal satellite measurements since 1978, which have been collated in a new volcanic SO2 emissions database (Carn, 2015c). This review discusses both the volcanological and atmospheric applications of the SO2 emissions database.
We do not offer a detailed overview of any specific volcanic eruptions, but instead synthesize
~36 years of satellite observations and augment previous compilations of satellite SO2 data for volcanic eruptions (e.g., Bluth et al., 1993, 1997; Carn et al., 2003) and estimates of global volcanic SO2 emissionsACCEPTED (e.g., Andres and Kasgnoc MANUSCRIPT, 1998; Halmer et al., 2002), incorporating a significant amount of new data on smaller eruptions and passive degassing collected since 2004.
The paper is organized as follows. Firstly (section 2), we briefly review recent developments in satellite remote sensing of volcanic gases, focusing on improvements in instrument sensitivity and in techniques to determine volcanic plume altitude. In section 3 we describe the ultraviolet
(UV), infrared (IR) and microwave satellite datasets used to generate the multi-decadal volcanic
SO2 emissions database. The main results are presented in section 4, including the temporal and
6 ACCEPTED MANUSCRIPT
latitudinal distribution of volcanic SO2 emissions, updated estimates of the annual volcanic SO2 flux from volcanic eruptions (including the fraction not recorded by satellite observations) and a reassessment of the relationship between eruptive SO2 discharge and eruption magnitude (VEI).
We also provide a detailed breakdown of the sources of the measured SO2 emissions and attempt to classify them based on degassing style. New constraints on e-folding times for SO2 removal in volcanic plumes are discussed, and we also highlight recent measurements of volcanic HCl injections into the UTLS. An analysis of the major sources of passive volcanic emissions of SO2 detected in satellite data is presented in section 4.2, followed by a discussion (section 5) and conclusions.
2. Satellite measurements of volcanic gases
Although to date no satellite instrument has been deployed to specifically target volcanic emissions, passive UV, IR and microwave sensors have been used to measure seven volcanic gas species since 1978 (SO2, H2S, HCl, CO, BrO, OClO and CH3Cl; Table 1). However, SO2 is by far the most frequently measured volcanic gas, and it is currently not possible to measure other volcanic gases from space with comparable accuracy or frequency to that of SO2. It is the major target for remote sensingACCEPTED due to its ease of measurement MANUSCRIPT relative to the dominant volcanic volatile species, which are typically water vapor (H2O) and carbon dioxide (CO2). The characteristics of SO2 that enable space-based measurements over at least 5 orders of magnitude in atmospheric column amount (without spatial or temporal data averaging) are its relative
nd rd abundance in volcanic emissions (typically 2 or 3 after H2O and CO2; e.g., Oppenheimer et al., 2011), low background SO2 concentrations away from strong anthropogenic point sources
(power plants and metal smelters; Carn et al., 2007), and the accessibility of absorption bands
7 ACCEPTED MANUSCRIPT
for remote sensing in the UV, IR and microwave spectral regions. The relatively high emission altitude (cf. anthropogenic sources) and consequent atmospheric lifetime of volcanic SO2
(typically hours to days) is also conducive to satellite measurements, the latter being sufficiently long to enable detection and tracking of volcanic SO2 plumes despite the large spatial scale of satellite sensor footprints, but short enough to prevent significant accumulation in the atmosphere and interference with fresh emissions (in contrast to CO2). To date, the only point sources of anthropogenic SO2 emissions observed to rival volcanic SO2 plumes in terms of signal strength have been metal smelters (e.g., Khokhar et al., 2005; Carn et al., 2007; Walter et al.,
2012; Baudin et al., 2014) and rare sulfur fires (Carn et al., 2004; Kearney et al., 2009), although, on a global scale, estimated total anthropogenic sulfur emissions may exceed volcanic sulfur emissions by up to an order of magnitude, with significant uncertainty (e.g., Chin and
Jacob, 1996; Graf et al., 1997; Andres and Kasgnoc, 1998). The main impediment to detection of volcanic CO2 and H2O emissions from space is the small volcanic contributions relative to the ambient atmospheric concentrations of these molecules. Whilst the detection of volcanic CO2 from space has not yet been conclusively demonstrated, it is likely to occur as more data from instruments providing CO2 retrievals, including GOSAT, OCO-2, SCIAMACHY, AIRS and
IASI (Table 1), are analyzedACCEPTED (e.g., Schwandner MANUSCRIPT and Carn, 2011; Schwandner et al., 2015). But, although several operational sensors retrieve profiles of atmospheric water vapor (Table 1), satellite detection of volcanic H2O is extremely unlikely due to even greater background interference than for CO2, with the possible exception of volcanic H2O injections into the dry stratosphere.
Satellite remote sensing of SO2 exploits the molecule‘s absorption bands at UV, IR and microwave wavelengths. Somewhat fortuitously, the overlap of SO2 absorption features in the
8 ACCEPTED MANUSCRIPT
UV and IR with spectral regions used to retrieve higher priority measurements or atmospheric constituents (e.g., ozone, water vapor, clouds and surface temperature) has ensured that most operational satellite instruments have included channels suitable for SO2 detection. UV measurements utilize SO2 absorption bands at wavelengths of 300-340 nm, in the Huggins ozone absorption band; hence ozone and SO2 must be retrieved simultaneously and large column amounts of ozone can impact the sensitivity of UV measurements to lower tropospheric or boundary layer SO2 (e.g., Krotkov et al., 2008). IR measurements of SO2 can exploit three vibrational absorption bands in the thermal IR at ~4 µm (ν1+ν3-band), ~7.3 µm (ν3-band) and
~8.6 µm (ν1-band). Of these, the 7.3 µm band is most often used since it is the stronger band
(providing greater sensitivity), is less susceptible to interference by volcanic ash (e.g., Corradini et al., 2009), and provides sensitive measurements of SO2 in the upper troposphere and above
(e.g., Rinsland et al., 1995; Prata et al., 2003; Carn et al., 2005; Prata and Bernardo, 2007;
Clarisse et al., 2008, 2012; Doeringer et al., 2012; Höpfner et al., 2013). It lacks sensitivity to lower tropospheric SO2 (<3-5 km altitude; i.e., small eruptions and passive degassing) notwithstanding relatively rare coincidences of surface temperature inversions, extreme pollution levels and low humidity (e.g., Boynard et al., 2014; Bauduin et al., 2014). The typically poor sensitivity to lower troposphericACCEPTED SO2 at 7.3 µm MANUSCRIPT is due to interference from water vapor absorption, but this also simplifies retrievals somewhat since the background 7.3 µm radiation originates from relatively homogeneous (on the scale of a volcanic cloud) lower tropospheric water vapor. The 4 µm and 8.6 µm bands are weaker but are situated in atmospheric windows, allowing retrieval of lower tropospheric SO2, but with the added complexity of dependence on variable background emissivity, especially over land (e.g., Realmuto et al., 1994, 1997; Teggi et al., 1999; Realmuto, 2000; Watson et al., 2004; Prata and Bernardo, 2007; Prata and
9 ACCEPTED MANUSCRIPT
Kerkmann, 2007; Kearney et al., 2008), atmospheric water vapor (Realmuto and Worden, 2000), daytime solar reflection at 4 µm, and interference from volcanic ash at 8.6 µm (Corradini et al.,
2009). Microwave limb sounding of SO2 in the UTLS by the Microwave Limb Sounder (MLS) instruments has exploited rotational absorption bands of the molecule at frequencies of 200-240
GHz (Read et al., 1993; Pumphrey et al., 2015).
The era of satellite observations of volcanic degassing (primarily SO2) can be broadly divided based on sensor specifications (mainly spectral and spatial resolution), which directly impact sensitivity to SO2 abundance and altitude and hence to different styles of volcanic activity. Between October 1978 and December 2005, four TOMS missions provided near- continuous multi-spectral, global, nadir UV observations at low spatial resolution (39-50 km at nadir), with occasional small data gaps and a major 19-month gap in 1995-96 between the
Meteor-3 and Earth Probe (EP) TOMS missions (Krueger et al., 1995, 2000; Fig. 1). The limited number of TOMS wavelengths restricted measurements to larger volcanic eruptions and extreme levels of passive SO2 degassing (e.g., Bluth et al., 1993; Carn et al., 2003; Carn, 2004). SBUV
(flown since 1978; Table 1) has also provided nadir UV observations of volcanic SO2 clouds, albeit with poorer spatial sampling than TOMS, but only for the largest eruptions (McPeters et al., 1984; McPeters, ACCEPTED1993; Schnetzler et al. 1995) MANUSCRIPT. Coincident with TOMS, the High-resolution
Infrared Radiation Sounder and subsequent iterations (HIRS, HIRS/2, HIRS/3, etc.), flown as part of the TIROS Operational Vertical Sounder (TOVS) package on NOAA polar-orbiters since
October 1978, provide multi-spectral, global IR SO2 measurements at 7.3 µm (Prata et al.,
2003). The HIRS SO2 data provide coverage of the 1995-96 TOMS data gap and also offer nighttime and high-latitude winter measurements, unlike TOMS.
10 ACCEPTED MANUSCRIPT
Hyperspectral UV measurements, offering increased sensitivity to lower tropospheric
SO2 due to optimized spectral resolution, began with the Global Ozone Monitoring Experiment
(GOME) in July 1995 (Burrows et al., 1999). GOME provided some of the first satellite observations of lower tropospheric SO2 from volcanic and anthropogenic sources despite its low spatial resolution (40×320 km at nadir; Eisinger and Burrows, 1998), and detected numerous volcanic eruptions during its operational lifetime (1995-2003; e.g., Afe et al., 2004; Thomas et al., 2005; Khokhar et al., 2006). However, a GOME inventory of volcanic SO2 mass loading has not been widely reported to date and hence GOME SO2 data are not included in the database presented here (Fig. 1). GOME was superseded by the Scanning Imaging Absorption
Spectrometer for Atmospheric Chartography (SCIAMACHY; Bovensmann et al., 1999), launched on Envisat in March 2002 and operational from 2002-2012. SCIAMACHY provided increased sensitivity to tropospheric SO2 over GOME due to higher spatial resolution (30×60 km at nadir), and also detected many volcanic SO2 emissions (e.g., Afe et al., 2004; Lee et al., 2008,
2009; Penning de Vries et al., 2014) and provided the first satellite observation of volcanic OClO
(Theys et al., 2014), but its spatial coverage was limited by alternation between nadir and limb mode observations. Thus, SCIAMACHY-derived volcanic SO2 mass loadings have also not been widely reported and ACCEPTEDare not used herein, although MANUSCRIPT there is potential for further analysis of the
SCIAMACHY limb mode data archive to constrain volcanic plume altitudes (e.g, Penning de
Vries et al., 2014). The Ozone Monitoring Instrument (OMI), launched on NASA‘s Aura satellite in July 2004, has provided operational, hyperspectral nadir UV SO2 measurements with a nadir spatial resolution of 13×24 km since September 2004 (Fig. 1; Krotkov et al., 2006; Yang et al., 2007). OMI SO2 measurements have been the primary source of data for continuation of the TOMS record of volcanic SO2 emissions, but with increased sensitivity to small eruptions
11 ACCEPTED MANUSCRIPT
and passive degassing (Carn et al., 2008b; Carn and Prata, 2010; McCormick et al., 2012, 2013;
Campion et al., 2012; Carn et al., 2013). Overlap between EP/TOMS in OMI in 2004-2005 provides opportunities for comparison of multi-spectral and hyperspectral SO2 retrievals for several eruptions (Fig. 1; Table 2), to assess continuity between satellite missions. Second generation GOME (GOME-2) instruments were launched on the MetOp-A and MetOp-B satellites in October 2006 and September 2012, respectively, offering higher spatial resolution than GOME (80×40 km) and high sensitivity to volcanic SO2 and BrO emissions (e.g., Theys et al., 2009; Nowlan et al., 2011; Rix et al., 2012; Theys et al., 2013), but lower sensitivity to passive SO2 degassing than OMI due to OMI‘s higher spatial resolution. However, as for GOME and SCIAMACHY, we have not made extensive use of GOME-2 data in this study. A further hyperspectral UV sensor, the Ozone Mapping and Profiler Suite (OMPS), was launched on the
Suomi National Polar-orbiting Partnership (SNPP) satellite in October 2011 and will be flown on future NOAA Joint Polar Satellite System (JPSS) polar-orbiters beginning in 2017. OMPS has variable spatial resolution (50×50 km in standard mode and 10×10 km in zoom mode, currently used once weekly on Saturdays) and comparable sensitivity to volcanic SO2 to SCIAMACHY,
OMI and GOME-2 (e.g., Yang et al., 2013) and is also being used to extend the UV volcanic SO2 measurements in conjunctionACCEPTED with OMI (Carn MANUSCRIPT et al. 2015; Fig. 1).
IR satellite sensors have shown a similar evolution to UV instruments. Following on from
TOVS/HIRS, multi-spectral TIR observations from the Moderate resolution Imaging
Spectroradiometer (MODIS) and Advanced Spaceborne Thermal Emission and Reflection radiometer (ASTER) began following the launch of Terra in December 1999, and a second
MODIS instrument was deployed on Aqua in May 2002. Both Terra and Aqua MODIS and
ASTER continue to operate at the time of writing. MODIS and ASTER data can be used for IR
12 ACCEPTED MANUSCRIPT
SO2 retrievals at moderate-to-high spatial resolution (90 m to 1 km) using channels at 7.3 µm
(MODIS) or 8.6 µm (MODIS and ASTER); these sensors have typically been used for specific volcanic eruption or degassing case studies that exploit their higher spatial resolution rather than the production of long-term volcanic SO2 emissions inventories (e.g., Rose et al., 2003; Watson et al., 2004; Urai, 2004; Pugnaghi et al., 2006; Kearney et al., 2008; McCarthy et al., 2008;
Matiella Novak et al., 2008; Corradini et al., 2009, 2010; Campion et al., 2010, 2012; Henney et al., 2012). This is largely due to the difficulty of processing the larger data volumes associated with high spatial resolution data, the limited spatial coverage of ASTER (a result of the trade-off between spatial resolution and swath width), plus the aforementioned interference from volcanic ash at 8.6 µm. Although high-resolution IR spectra were acquired from space as early as 1969
(Clarisse et al., 2011a), global, nadir hyperspectral IR measurements began from orbit in earnest with the launch of the Atmospheric Infrared Sounder (AIRS) on Aqua in 2002 (Chahine et al.,
2006), which provides robust retrievals of volcanic SO2 in the UTLS using the 7.3 µm band
(Carn et al., 2005; Prata and Bernardo, 2007; Prata et al., 2007, 2010). AIRS spectra also cover the 4 µm band and part of the 8.6 µm SO2 band but these data have not been widely exploited to date (Prata and Bernardo, 2007). AIRS SO2 measurements provide particularly valuable observations of volcanicACCEPTED eruptions between 2002 MANUSCRIPT and the beginning of OMI measurements in
September 2004 (Fig. 1), in addition to ongoing complementary coverage of nighttime and high- latitude winter events. Hyperspectral IR measurements of SO2 were augmented by the Infrared
Atmospheric Sounding Interferometer (IASI), first launched on MetOp-A with GOME-2 in 2006 and followed by a second IASI instrument on MetOp-B in 2012. In addition to providing high- quality UTLS volcanic SO2 measurements (e.g., Clarisse et al., 2008, 2012, 2014; Karagulian et al., 2010; Theys et al., 2013), IASI has provided new observations of other volcanogenic gas and
13 ACCEPTED MANUSCRIPT
aerosol species including hydrogen sulfide (H2S; Clarisse et al., 2011b) and sulfate aerosol
(Clarisse et al., 2010; Karagulian et al., 2010).
In tandem with the advent of hyperspectral measurements, a key advance in satellite remote sensing that has occurred in the past decade, and one that is particularly relevant to volcanic eruption detection, is the availability of altitude information for SO2 (and other trace gases) and aerosols to complement measurements of SO2 mass loading by the nadir UV and IR imagers and sounders describe above. Although it has long been possible to estimate the altitude of a volcanic cloud by comparing observed SO2 or ash cloud transport and winds aloft (e.g., from a radiosonde), by using IR cloud-top brightness temperatures for optically thick plumes, or by using trajectory models (e.g., Schoeberl et al., 1993; Carn et al., 2008; Krotkov et al., 2010), there are now multiple sources of altitude information, with variable accuracy. Altitude information can be derived directly from satellite instruments in several ways including: active
LiDAR or radar remote sensing of volcanic aerosols and hydrometeors (Fig. 2; e.g., Cloud-
Aerosol Lidar with Orthogonal Polarization (CALIOP) or CloudSat data; Carn et al., 2009a,
2015; Spinei et al., 2010; Carn and Lopez, 2011; Winker et al., 2012; Fromm et al., 2014; Prata et al., 2015); passive multi-angle imaging (e.g., the Multi-angle Imaging Spectroradiometer
[MISR]; Kahn et al.,ACCEPTED 2007; Kahn and Limbacher MANUSCRIPT, 2012); passive limb sounding of volcanic gases and aerosols in the UV, IR or microwave (e.g., MLS, OSIRIS, MIPAS, SCIAMACHY,
ACE, OMPS; Read et al., 1993; Bourassa et al., 2010, 2012, 2013; Doeringer et al., 2012;
Höpfner et al., 2013, 2015; Penning de Vries et al., 2014; Pumphrey et al., 2015); solar occultation measurements of volcanic aerosols (e.g., SAGE); or SO2 altitude retrievals using passive nadir hyperspectral UV or IR measurements (e.g., OMI, GOME-2, IASI; Yang et al.,
2009, 2010; Nowlan et al., 2011; Rix et al., 2012; Clarisse et al., 2014). Information on the SO2
14 ACCEPTED MANUSCRIPT
vertical profile can also be gleaned from inverse trajectory modeling of satellite measurements
(e.g., Eckhardt et al., 2008; Kristiansen et al., 2010; Hughes et al., 2012) or by comparisons of simultaneous UV and IR measurements, exploiting their different vertical sensitivity (e.g., Fig. 2;
Carn et al., 2009b). Whilst it is rare to possess all the above information for a given volcanic eruption, the redundancy offered by multiple techniques means that some altitude information is nearly always available. The operation of polar-orbiting satellites in close formation along the same orbit track, such as NASA‘s A-Train afternoon constellation (http://atrain.nasa.gov/; including OCO-2, Aqua, CALIPSO, CloudSat and Aura), providing near-coincident passive and active measurements from multiple platforms, has greatly facilitated analysis of volcanic cloud altitude (e.g., Fig. 2; Carn et al., 2009a, 2009b).
To summarize, there have been many improvements in satellite remote sensing of volcanic gases since 1978, largely due to the advent of hyperspectral UV and IR sensors and satellite constellations. Currently (June 2015), as several operational sensors continue to exceed their intended lifetimes, our capabilities for monitoring volcanic SO2 emissions from space have reached a zenith, with up to ~18 daily SO2 observations potentially available from nadir imagers and sounders. In chronological order, these are: AIRS and MODIS (Aqua; equator crossing at
1:30 am local time (LT))ACCEPTED; VIIRS and CrIS (SNPP; MANUSCRIPT 1:30 am LT), GOME-2B and IASI-B
(MetOp-B; 9:00 am LT); GOME-2A and IASI-A (MetOp-A; 9:30 am LT); MODIS (Terra;
10:30 am LT); AIRS and MODIS (Aqua; 1:30 pm LT); OMPS, VIIRS and CrIS (SNPP; 1:30 pm
LT); OMI (Aura; 1:45 pm LT); IASI-B (9:00 pm LT); IASI-A (9:30 pm LT); and MODIS
(Terra; 10:30 pm LT). This sequence does not include further observations from multiple consecutive overpasses for high latitude regions (e.g., Kamchatka, Iceland, Alaska), daily SO2 observations from TOVS/HIRS, limb observations of SO2 and HCl from MLS on Aura and
15 ACCEPTED MANUSCRIPT
occasional SO2 observations by ASTER and the geostationary SEVIRI (for eruptions in Europe,
Africa and the eastern Americas). Note that the times listed above are equatorial crossing times
(ascending or descending nodes) and the gap between local satellite overpass times may be substantially longer at higher latitudes, although this is offset by the convergence of polar orbits towards the poles. But at low latitudes the maximum data gap between volcanic SO2 measurements (UV or IR) is currently ~7.5 hours (1:30 am – 9:00 am and 1:30 pm – 9:00 pm), which is useful for near real-time (NRT) monitoring applications (e.g., Brenot et al., 2014) and ensures that very few, if any, volcanic eruptions emitting significant SO2 go undetected. Having numerous satellite SO2 measurements available also permits inter-comparisons between UV and
IR data and increases confidence in calculated SO2 mass loadings. Furthermore, the recently launched Earth Polychromatic Imaging Camera (EPIC) on the NOAA Deep Space Climate
Observatory (DSCOVR) will soon provide the first volcanic SO2 measurements from the L1
Earth-Sun Lagrange point (http://www.nesdis.noaa.gov/DSCOVR/). Beginning in 2016, EPIC will provide multi-spectral UV SO2 measurements of the sunlit Earth disk several times per day, with spatial resolution similar to OMI, offering improved temporal resolution over LEO UV sensors. The L1 orbit offers a unique perspective of Earth, distinct from a geostationary orbit
(GEO). Like GEO, yearACCEPTED-round L1 observations MANUSCRIPT will be best suited to eruptions at low- to mid- latitudes, but whereas GEO observations are always limited at high latitudes, EPIC will provide good coverage of the high latitude summer hemisphere as the Earth‘s axis tilts towards the Sun.
3. Satellite instruments and data
Here, we describe the specific datasets used in the multi-decadal analysis of volcanic SO2 emissions in chronological order of sensor deployment, namely TOMS, TOVS/HIRS, AIRS,
16 ACCEPTED MANUSCRIPT
OMI, MLS, IASI and OMPS. We do not provide detailed descriptions of retrieval algorithms, since these have been extensively reported elsewhere and the focus of this review is interpretation of the measurements. Limitations and uncertainties in SO2 retrievals will be acknowledged as appropriate, but we will also demonstrate broad agreement between SO2 retrievals from independent UV and IR sensors (especially for the larger volcanic eruptions), providing confidence in the measurements. However, we also note that retrieval algorithms are likely to improve in the future due to continued improvements in spectroscopic data, additional constraints from ancillary data and better characterization of uncertainties. Thomas and Watson
(2009) and Carn (2015a) provide recent tabulations of satellite sensor specifications.
3.1 TOMS (UV)
A TOMS instrument was first deployed on the Nimbus-7 (N7) satellite that was operational between October 1978 and May 1993, and N7/TOMS provided the first satellite observation of volcanic SO2 after the 1982 El Chichón eruption (Krueger, 1983; Kruger et al., 2008).
Subsequently, three similar 6-channel TOMS sensors were launched on the Meteor-3 (M3;
August 1991 – December 1994), ADEOS (AD; September 1996 - June 1997) and Earth Probe
(EP; July 1996 – DecemberACCEPTED 2005) spacecraft. FollowingMANUSCRIPT the 1982 El Chichón SO2 cloud sighting, algorithms were developed for quantitative retrievals of SO2 column amounts (Krueger et al., 1995; 2000). The N7, M3, AD and EP/TOMS instruments differed in their channel placement (after discovery of the volcanic SO2 mapping capability, AD/TOMS and EP/TOMS wavelengths were optimized for SO2 measurements; Gurevich and Krueger, 1997) and spatial resolution (24-50 km nadir footprint size), which impacts SO2 sensitivity: the estimated volcanic
SO2 cloud detection limit ranges from ~2-4 kt (EP/TOMS) to ~11 kt (N7/TOMS; Carn et al.,
17 ACCEPTED MANUSCRIPT
2003). There was a TOMS data gap of ~19 months in 1995-96 between the M3 and EP/TOMS missions, but we use IR TOVS SO2 measurements to cover this time period (Prata et al., 2003).
The Solar Backscatter Ultraviolet (SBUV or SBUV/2) instruments flown on the N7 satellite with
TOMS and later on NOAA polar-orbiting satellites could also detect volcanic SO2 (e.g.,
McPeters et al., 1984; McPeters, 1993; Schnetzler et al., 1995), but with limited mapping ability
(Table 1), and to date have only been applied to the largest eruptions (El Chichón and Pinatubo).
Most TOMS volcanic SO2 data reported here have been produced using the iterative 4- band matrix inversion retrieval technique described by Krueger et al. (1995, 2000), using Sun- normalized radiances produced by the TOMS ozone data production algorithm as input, and effective optical paths derived from a UV radiative transfer model. Krueger et al. (1995) report errors of ±10-30% for TOMS SO2 retrievals, highest at large solar zenith angles near the swath edge. Previous TOMS volcanic SO2 analyses used TOMS production data version 6 (e.g., Bluth et al., 1992, 1993, 1997) or version 7 (Carn et al., 2003), whereas more recent analyses have used version 8 (the current TOMS ozone algorithm is version 9). Some, though not all, older eruptions have been reanalyzed using version 8 TOMS data, with little change from previously reported results (e.g., Krueger et al., 2008). TOMS production algorithm updates are largely focused on improvingACCEPTED the accuracy of total ozone MANUSCRIPT retrievals by correcting for the effects of aerosols, clouds and ozone profile variability (e.g., Bhartia and Wellemeyer, 2002), and hence small improvement in SO2 retrievals may also result. TOMS SO2 column retrievals are not publicly available at present, but will be released following a major reprocessing of the TOMS data archive that is expected to be complete by 2018 (in the meantime, TOMS SO2 retrievals are available from the first author upon request). Currently, N7/TOMS and EP/TOMS level 2 (L2) ozone data (version 8) are publicly available from the NASA Goddard Earth Sciences (GES)
18 ACCEPTED MANUSCRIPT
Data and Information Services Center (DISC; http://disc.sci.gsfc.nasa.gov/), and the L2 TOMS ozone data files contain a SO2 index (SOI) indicating contamination of ozone retrievals by volcanic SO2.
As is the case for all satellite-based volcanic SO2 measurements, opportunities to validate
TOMS SO2 data were rare and opportunistic, relying on the chance overpass of a volcanic cloud over a ground measurement site. The best validation data have been obtained using ground-based
UV Brewer spectrophotometers, which can be used to measure total column SO2 and are operated as networks in several countries (e.g., Fioletov et al., 1998). A Brewer instrument in
Toronto (Canada) detected SO2 (column amounts up to ~100 DU) in the May 1980 Mt. St.
Helens (WA, USA) eruption cloud as it drifted east, with general agreement between TOMS and
Brewer SO2 columns (Kerr and Evans, 1987). In 1984, a Brewer in Sweden detected SO2 (>40
DU) emitted by the September 5, 1984 effusive eruption of Krafla (Iceland); in this case significant disagreement between Brewer and TOMS SO2 observations was attributed to lower tropospheric SO2 which was not detected by TOMS (Kerr and Evans, 1987). Perhaps the most robust TOMS SO2 validation occurred when the Toronto Brewer measured SO2 in a drifting volcanic cloud from Mt. Spurr (Alaska) in September 1992, with coincident N7/TOMS and
Brewer SO2 columnsACCEPTED (~40-50 DU) agreeing to MANUSCRIPT within 20% (Fioletov et al., 1998; Krueger et al.,
2000). Attempts to compare TOMS SO2 mass measurements with ground-based UV correlation spectrometer (COSPEC) SO2 emission rates were made for the first time during the 1982
Galunggung eruption (Java, Indonesia; Bluth et al., 1994). In this case TOMS and COSPEC observed different emissions (eruptive and non-eruptive, respectively), precluding validation but demonstrating the complementarity of the satellite and ground-based measurements.
19 ACCEPTED MANUSCRIPT
3.2 TOVS/HIRS (IR)
The High-resolution Infrared Radiation Sounder (HIRS/2) was first launched as part of the
TOVS package on the TIROS-N satellite in October 1978, and subsequent iterations of the HIRS instrument (HIRS/3, HIRS/4) have been flown on all NOAA polar-orbiters and more recently on the European MetOp satellites, providing continuous coverage to the present. Prata et al. (2003) provide further details on TOVS/HIRS and describe a technique for retrieving UTLS SO2 column amounts from HIRS/2 L1b radiance data, which we use to analyze volcanic eruptions that occurred during the 1995-96 TOMS data gap. However, the full HIRS/2 SO2 dataset could potentially be used to generate a long-term volcanic SO2 climatology for 1978-present, with greater temporal sampling than UV sensors, but reduced sensitivity to tropospheric SO2. Prata et al. (2003) and Carn et al. (2008a) show good agreement between TOMS and HIRS/2 SO2 retrievals for a small subset of eruptions that have been analyzed to date with both instruments
(e.g., Table 2).
3.3 AIRS (IR)
AIRS, in Sun-synchronous orbit on Aqua in the A-Train, provides hyperspectral IR measurements using ACCEPTED2378 channels at IR wavel MANUSCRIPTengths between ~3.3-16.7 µm (Chahine et al.,
2006). This wavelength range covers SO2 absorption bands at 4 and 7.3 µm, and part of the 8.6
µm band (there is a gap from ~8.2-8.8 µm). The primary goal of AIRS is to provide global atmospheric soundings of temperature, moisture and other gases for numerical weather prediction and climate, but Carn et al. (2005) and Prata and Bernardo (2007) have shown that
AIRS can be used to retrieve volcanic SO2 at altitudes above ~3-5 km using the 7.3 µm band.
AIRS scans a swath of ±49º from nadir and the sensor‘s elliptical footprint dimensions vary from
20 ACCEPTED MANUSCRIPT
15×15 km at nadir to 18×40 km at the swath edge. A full description of the AIRS instrument and operational data processing can be found at http://airs.jpl.nasa.gov.
IR AIRS SO2 retrievals used here are derived using the Level 1B AIRS IR geolocated and calibrated radiance product (AIRIBRAD version 5) and the Prata and Bernardo (2007) algorithm. AIRIBRAD data granules are publicly available from the NASA GES DISC
(http://disc.sci.gsfc.nasa.gov/). The Prata and Bernardo (2007) algorithm is also used to produce
NRT AIRS SO2 products for aviation hazard mitigation, which can be obtained via the NASA
GES DISC (http://disc.sci.gsfc.nasa.gov/nrt/data-holdings/airs-nrt-products) and viewed on the
Support to Aviation Control Service (SACS) website (http://sacs.aeronomie.be/nrt; Brenot et al.,
2014). Prata and Bernardo (2007) quote a retrieval accuracy (or detection limit) of 6 DU SO2
(~1.6×1017 molecules cm-2) for a single AIRS pixel retrieval, mostly due to errors in accounting for background atmospheric water vapor and clouds. Thomas et al. (2011) report a similar AIRS
SO2 detection limit of 5 DU. A detection limit of 6 DU equates to ~30-100 tons of SO2 within a single AIRS footprint, depending on swath position, and assuming that a cluster of ~10 pixels would constitute a statistically significant volcanic cloud yields a minimum detectable SO2 mass of ~0.3-1 kilotons (kt) in the UTLS. In practice, minimum reported volcanic SO2 mass loadings derived from AIRS (forACCEPTED SO2 clouds consisting MANUSCRIPTof multiple pixels) are ~0.3-10 kt depending on eruption latitude (e.g., Prata and Bernardo, 2007; Carn and Prata, 2010; Bonny et al., 2015), indicating slightly higher sensitivity to UTLS SO2 than EP/TOMS in the 2002-2004 period of overlap between AIRS and TOMS. We also exploit the synergy between near-coincident AIRS
7.3 µm partial column SO2 retrievals and OMI total column retrievals from the A-Train to infer volcanic SO2 altitude, where OMI≈AIRS>0 implies UTLS SO2, OMI>AIRS>0 implies a vertically extensive SO2 cloud straddling the lower troposphere and UTLS (or alternatively, a
21 ACCEPTED MANUSCRIPT
multi-layered volcanic cloud with layers residing at multiple altitudes), and OMI>AIRS≈0 implies SO2 confined to the lower troposphere (e.g., Carn et al., 2009b).
To date, efforts to validate AIRS SO2 measurements have focused on inter-comparisons with UV SO2 retrievals from TOMS, OMI and OMPS. Good correlations have been found between AIRS and UV SO2 burdens for volcanic clouds at high altitude (>8 km; e.g., Fig. 3;
Prata and Bernardo, 2007; Carn and Prata, 2010; Thomas et al., 2011; Carn et al., 2015). The weak sensitivity of AIRS to lower tropospheric SO2, along with incomplete coverage by the
AIRS swath for some eruptions, results in a poorer correlation for eruptions to lower altitudes
(Prata and Bernardo, 2007). As yet there have been few detailed comparisons between IR and
UV SO2 column retrievals at the pixel scale, principally due to problems with spatial and temporal collocation and typically non-uniform SO2 distributions in volcanic clouds (e.g.,
Kearney et al., 2009; Carn et al., 2015).
3.4 OMI (UV)
OMI is a hyperspectral UV-Visible spectrometer (270-500 nm) using a two-dimensional charge- coupled device (CCD) detector to collect simultaneous spectral and spatial measurements of backscattered solar radiationACCEPTED across a 2600 km MANUSCRIPT swath (Levelt et al., 2006). Spatial resolution is
13×24 km (along-track × across-track) at nadir, increasing to ~28×160 km at the swath edge. The
OMI UV-2 channel (307-383 nm) used for SO2 retrievals has an average spectral resolution of
0.45 nm (Levelt et al., 2006a), and this high spectral resolution coupled with the small OMI footprint size provides unprecedented SO2 sensitivity for a satellite instrument, permitting detection of passive volcanic degassing and anthropogenic pollution (e.g., Carn et al., 2007,
2008b, 2013). OMI became fully operational in September 2004 and provided daily contiguous,
22 ACCEPTED MANUSCRIPT
global daytime mapping of ozone, SO2 and other trace gases (NO2, BrO, HCHO) until May
2008, when a blockage appeared in the sensor‘s field of view. This ‗row anomaly‘ evolved dynamically (generally worsening) over time to obscure a significant proportion of the OMI swath from nadir to the east (see: http://www.knmi.nl/omi/research/product/rowanomaly- background.php), ending daily contiguous coverage at low latitudes, although data unaffected by the anomaly remain of high quality. Future evolution of the row anomaly is hard to predict.
Several algorithms have been used to retrieve SO2 from OMI UV radiances, with some differences in SO2 column sensitivity and retrieval dynamic range (e.g., Krotkov et al., 2006;
Yang et al., 2007, 2009a, 2009b, 2010; Lee et al., 2009; Li et al., 2013; Carn et al., 2013; Theys et al., 2015). Most of the volcanic SO2 measurements in the current inventory are derived from the Linear Fit (LF) algorithm (Yang et al., 2007), which has been used to generate the operational, publicly available OMI SO2 product (OMSO2 collection 3; available from the
NASA GES DISC). The LF algorithm uses 10 OMI wavelengths at 310-360 nm, covering regions of strong (shorter wavelengths) and weak (longer wavelengths) SO2 absorption, to retrieve SO2 column amounts (Yang et al., 2007). By using longer UV wavelengths, the LF algorithm avoids some of the non-linear saturation effects that limit algorithms using shorter wavelengths to lowerACCEPTED SO2 column amounts (~10 MANUSCRIPT-20 DU; Krotkov et al., 2006). However, the use of a zero SO2 a-priori constraint as a linearization point also limits the LF algorithm to column amounts of ~100 DU (Yang et al., 2007), and hence offline iterative algorithms have been developed to retrieve higher SO2 columns (Yang et al., 2009a). Such high SO2 column amounts
(>100 DU) are typically only encountered in very fresh volcanic eruption clouds (<1 day old), after which SO2 columns are rapidly depleted by upper atmospheric wind shear (e.g., Krueger et al., 2008), or close to the vent during SO2-rich effusive eruptions (e.g., Yang et al., 2009a). With
23 ACCEPTED MANUSCRIPT
the exception of highly concentrated SO2 plumes from effusive eruptions (>1000 DU; Yang et al., 2009a), the highest SO2 columns detected from space in UTLS volcanic clouds to date were measured by N7/TOMS ~6 hours after the April 4, 1982 El Chichón eruption (~600 DU;
Krueger et al., 2008) and ~21 hours after the paroxysmal June 15, 1991 eruption of Pinatubo
(~540 DU; Guo et al., 2004). No explosive eruption comparable in SO2 release to El Chichón or
Pinatubo has occurred during the OMI mission, but iterative SO2 retrieval techniques (e.g., Yang et al., 2009a) would certainly be required for accurate SO2 mass estimation in such cases. For comparison, peak SO2 column amounts of 200-300 DU were measured by OMI in the August
2008 Kasatochi volcanic cloud, among the largest to have occurred since Pinatubo (Krotkov et al., 2010). In the current volcanic SO2 inventory, most eruptions have been analyzed using the operational LF algorithm, with several of the larger explosive eruptions (e.g., Okmok and
Kasatochi in 2008; Nabro and Grimsvötn in 2011) also analyzed using offline algorithms to check for SO2 column saturation. Other consistency checks, such as monitoring the temporal evolution of SO2 loading in the days following an eruption, checking coincident ozone fields for artifacts (anomalies in ozone retrievals indicate that SO2 absorption has been incorrectly accounted for), and comparing with IR SO2 retrievals (e.g., Fig. 3), are also routinely used.
SO2 retrieval ACCEPTEDalgorithms require specification MANUSCRIPT of SO2 altitude due to the temperature- and pressure-dependence of SO2 absorption and the air mass factor (the ratio of slant to vertical column density). In the LF algorithm, and others, this is achieved by retrieving separate SO2 column amounts for three prescribed SO2 vertical profiles representing different volcanic degassing styles (e.g., Yang et al., 2007; Rix et al., 2012; Carn et al., 2013; Theys et al., 2015).
The vertical profiles used are (Carn et al., 2013): SO2 in the lower troposphere (TRL; 0–5 km altitude with center of mass altitude (CMA) of 2.5 km); SO2 in the mid-troposphere (TRM; 5–10
24 ACCEPTED MANUSCRIPT
km altitude, CMA=7.5 km) and SO2 in the lower stratosphere (STL; 15–20 km altitude;
CMA=17 km). Corresponding styles of volcanic degassing are passive emissions in the lower troposphere (TRL), high-altitude volcanic degassing and/or small-to-moderate eruptions (TRM) and large eruptions with stratospheric injection (STL). Our approach for volcanic eruption analysis is to use available information on plume altitude (see section 2, and below) to select the appropriate SO2 CMA, or interpolate between CMAs for intermediate altitudes. Information on altitude can include direct SO2 altitude retrievals using OMI measurements (e.g., Yang et al.,
2009b, 2010). Provided that SO2 column saturation is avoided and the assumed SO2 vertical profile is approximately correct, LF SO2 retrievals have an uncertainty of ~20% (Yang et al.,
2007).
In addition to SO2 mass, we also use an OMI SO2 index (OSI) calculated from OMI ozone algorithm (OMTO3) residuals (Krotkov et al., 2006) to identify volcanic SO2 emissions
(e.g., Fig. 4). We use the OSI as a proxy for volcanic SO2 to avoid complications related to calculation of SO2 mass when SO2 from multiple sources at different altitudes may be present
(due to the difficulty of isolating SO2 at different altitudes), a common problem in UV measurements which are sensitive to SO2 anywhere in the atmospheric column in cloud-free conditions (e.g., YangACCEPTED et al., 2007). This is less MANUSCRIPT of an issue with IR SO2 retrievals using the 7.3
µm band since these effectively screen out any lower tropospheric SO2.
OMTO3 residuals are calculated using scaled, normalized reflectances (N-values) where
N-values are defined as:
I N 100log (1) F
25 ACCEPTED MANUSCRIPT
where I is the backscattered Earth radiance and F is the solar irradiance at wavelength λ. N-value residuals for a pair of wavelengths (resj) are computed thus:
measured calculated resj N j N j (2)
short long NNNj j ( j ) (3)
where calculated N-values are derived from a forward radiative transfer model, accounting for
Rayleigh scattering, ozone absorption, Ring effect (Raman scattering) and surface reflectivity, but not SO2 absorption. The subscript j (j = 1, 2, 3) denotes wavelength pairs where pairs 1, 2 and 3 (P1, P2 and P3) correspond to wavelengths of 311.9–310.8 nm, 313.2–311.9 nm and
314.4–313.2 nm, respectively. These wavelength pairs represent maxima and minima in the UV
SO2 absorption cross-section (Krotkov et al., 2006; Carn et al., 2013) and this wavelength placement permits detection of SO2 contamination (independent of SO2 column retrieval) if the following relationships are simultaneously satisfied (Carn et al., 2013):
res 0 ACCEPTED MANUSCRIPTP1 resP2 0 (4)
resP3 0
The OSI is defined using the N-value pair residuals as follows:
OSI resPPP1 abs res 2 res 3 (5)
26 ACCEPTED MANUSCRIPT
The value of the OSI thus increases with the severity of SO2 contamination (higher SO2 column amount and/or SO2 altitude), but since it uses short UV wavelengths it also saturates at a certain point for large eruptions. Here, we use the OSI as a semi-quantitative indication of volcanic SO2 emissions and also to detect sources of passive volcanic degassing. In the latter case saturation is not an issue due to the much lower SO2 column amounts and plume altitudes involved.
There have been several efforts to validate OMI SO2 retrievals, including inter- comparisons with IR satellite measurements (e.g., Carn and Prata, 2010), ground-based SO2 measurements (e.g., Spinei et al., 2010; Carn and Lopez, 2011; Theys et al., 2015; Ialongo et al.,
2015) and airborne in-situ measurements (e.g., Krotkov et al., 2008; Carn et al., 2011). Results suggest that OMI SO2 measurements are generally consistent with independent retrievals in dilute (aged) volcanic clouds, but validation of the very high SO2 column amounts present in fresh volcanic eruption clouds remains a challenge, mainly due to the absence of ground-based or in-situ data in such situations.
To detect sources of passive SO2 degassing using the OSI we have adopted the following procedure, which exploits the fact that the locations of all subaerial volcanoes (i.e., the sources of passive SO2 emissions)ACCEPTED are known and cataloged MANUSCRIPT in the VOTW database. Detection and attribution of small, non-eruptive volcanic SO2 plumes also requires sufficient spatial resolution to identify the source, which OMI provides in most cases although ambiguity can still arise in regions with several actively degassing volcanoes. For each day of global OMI SO2 measurements, all pixels with identifiable SO2 contamination (based on equation 4 above) are extracted from the OMSO2 data products and ranked based on their OSI values. In doing this, we attempt to exclude pixels affected by the South Atlantic radiation Anomaly (SAA) over South
27 ACCEPTED MANUSCRIPT
America (which have distinctive N-value residuals) and also exclude pixels with solar zenith angles >65º to reduce noise. Then, for each pixel we identify the closest active volcano within a
50 km radius of the pixel center coordinates using a modified version of the VOTW catalog, which is assigned as the presumed source of the detected SO2; the 50 km distance threshold is arbitrary, but further afield the source of detected SO2 becomes increasingly ambiguous. We acknowledge that there are uncertainties associated with this approach (e.g., the possible presence of more than one volcano within 50 km of the anomaly), but we attempt to mitigate these by cross-checking with available activity reports to exclude unlikely sources. Also, the modified volcano list used here includes the most likely potential sources of SO2 emissions based on historical activity, and is necessary to facilitate source attribution in some regions (note that this modification does not impact SO2 detection, only source identification). Once sources have been assigned to all the SO2 contaminated pixels, the unique sources are identified (eliminating multiple detections of the same SO2 plume; e.g., where multiple SO2 contaminated pixels occur within a SO2 plume from a single source volcano) and cataloged for each day. By repeating this process for the lifetime of the OMI mission we acquire an inventory of the dominant sources of passive SO2 emissions detected from space. The analysis presented here is based on OMI SO2 observations from theACCEPTED beginning of the mission MANUSCRIPT (6 September, 2004) to the end of 2014, or ~3700 days of SO2 measurements.
3.5 MLS (microwave) and IR limb sounders
The first MLS instrument was deployed on NASA‘s Upper Atmosphere Research Satellite
(UARS) in September 1991, shortly after the Pinatubo and Cerro Hudson eruptions, and continued to operate until August 2001 but with only intermittent measurements after December
28 ACCEPTED MANUSCRIPT
1993 (http://mls.jpl.nasa.gov/uars/). UARS MLS detected the tropical stratospheric SO2 veil produced by the Pinatubo eruption (Read et al., 1993) and the large April 1993 eruption of
Lascar volcano (Chile; http://mls.jpl.nasa.gov/joe/so2_lascar.html). UARS MLS data were used to constrain the altitude (~26 km), SO2 decay e-folding time (~29-41 days) and total SO2 mass
(~17 Tg) of the Pinatubo SO2 cloud (Read et al., 1993). However, UARS MLS data have not been analyzed in detail for this work. The second MLS was launched on Aura (with OMI) in July
2004 (Waters et al., 2006) and continues to operate at the time of writing. Aura/MLS measures microwave thermal emission from Earth‘s limb to retrieve vertical profiles of trace gas mixing ratios in the UTLS; its great value for volcanic gas detection is its ability to provide vertical profiles of SO2 and HCl (and also ice water content; e.g., Fig. 2) coincident with nadir OMI and
AIRS measurements from the A-Train. Hence we use daily Aura/MLS limb soundings of SO2 and HCl to pinpoint UTLS injections of these species since September 2004 (e.g., Fig. 2, 4;
Pumphrey et al., 2015). The main weakness of MLS is poor spatial sampling: profiles (with vertical resolution of ~3 km) are retrieved only along the sub-satellite Aura orbit track with horizontal spacing of ~167 km (Pumphrey et al., 2015). Hence smaller volcanic eruptions may not be sampled, but these are also less likely to reach the UTLS. We use version 3.3 of the level
2 MLS Daily SO2 andACCEPTED HCl products, collection MANUSCRIPT 3 (ML2SO2.003 and ML2HCL.003; Froidevaux et al., 2008). ML2SO2 and ML2HCL data products are available from the NASA GES DISC
(http://disc.sci.gsfc.nasa.gov/). Further details on MLS and SO2 retrieval procedures are given by
Pumphrey et al. (2015), and Livesey et al. (2011) provide information on the precision, accuracy and useful vertical range of MLS SO2 and HCl retrievals, which are 3 ppbv, 5-20 ppbv and 215-
10 hPa for SO2 and 0.2-0.6 ppbv, 0.2-0.4 ppbv and 100-0.32 hPa for HCl, respectively. We note,
29 ACCEPTED MANUSCRIPT
however, that the very strong perturbations to SO2 and HCl mixing ratios caused by volcanic eruptions can permit detection in MLS data at pressure levels outside the stated useful range.
In addition to MLS data, information on SO2 vertical profiles in volcanic clouds for some eruptions has also been supplied by other limb sounding instruments (Table 1) such as MIPAS
(2002-2012; Höpfner et al., 2013, 2015) and ACE (2003-present; Doeringer et al., 2012), and by
Aura/TES (a nadir hyperspectral sounder; Clerbaux et al., 2008). MIPAS and ACE are both
Fourier Transform Spectrometer (FTS)-based sensors operating in the TIR using the 7.3 µm SO2 band, but MIPAS detected limb thermal emission whilst ACE uses solar occultation twice daily
(sunrise and sunset). Both MIPAS and ACE data can suffer from poor spatial and temporal sampling (hence identification of the source of volcanic SO2 perturbations can be ambiguous; e.g., Brühl et al., 2015); for example, Höpfner et al. (2013) analyzed monthly mean MIPAS limb spectra binned within 10º-wide latitude bands in order to detect small amounts of SO2 in the upper stratosphere. Höpfner et al. (2015) analyzed single MIPAS limb scans and identified over
30 volcanic SO2 injections into the UTLS between 2002 and 2012, some of which were from unidentified sources. However, our database of global volcanic SO2 emissions based on nadir satellite observations can be used to pinpoint these sources: e.g., tropical UTLS injections of SO2 from an unidentifiedACCEPTED source detected by MIPAS MANUSCRIPT in July-August 2005 (Höpfner et al., 2015) can be confidently ascribed to eruptions of Anatahan (CNMI; Fig. 4).
3.6 IASI (IR)
IASI is a nadir-viewing Fourier-transform IR (FTIR) spectrometer first launched on the polar- orbiting MetOp-A satellite on 19 October 2006 (Clerbaux et al., 2007; Phulpin et al., 2007). A second, identical IASI instrument was launched on MetOp-B on 17 September 2012. Both IASI
30 ACCEPTED MANUSCRIPT
instruments provide global coverage twice a day, with a nadir footprint size of 12 km (diameter) and a full swath width of 2200 km. The MetOp satellites are operational platforms; hence the
IASI measurements are available in near real time. The IASI spectrometer provides continuous spectral coverage from 645 cm−1 to 2760 cm−1 (~3.6-15.5 µm) with no gaps (in contrast to
AIRS), with full coverage of the ν1, ν3 and ν1+ν3 SO2 absorption bands. IASI measurement noise is also lower than AIRS; the IASI noise-equivalent temperature change (NEΔT) at 280 K is estimated to be 0.05 K in the ν3 band and 0.12 K in the ν1 band, significantly better than the expected 0.2–0.35 K. For comparison, the AIRS NEΔT is ~0.2 K (but note that this is at a scene temperature of 250 K; at the warmer temperature of 280 K quoted for IASI the AIRS NEΔT will be higher) and it only partially covers the ν1 band (Carn et al., 2005).
IASI SO2 retrievals (Fig. 7, 8) employ the technique described by Clarisse et al. (2012), using an SO2 altitude retrieval algorithm (Clarisse et al., 2014) to derive accurate plume altitudes and hence more reliable estimates of SO2 mass loadings. Inaccurate SO2 loadings may still result if the plume altitude is incorrectly retrieved. Furthermore, as for most IR 7.3 µm SO2 retrievals, the IASI SO2 measurements are only reliable for SO2 at altitudes above ~5 km (in mid-latitudes) or ~7 km (in the tropics) due to water vapor interference. SO2 plumes at lower altitudes will be missed or mass loadingACCEPTED will be underestimated MANUSCRIPT (this could be mitigated in the future by exploitation of IASI data in the ν1 SO2 band). This is a pertinent issue for relatively long-lived eruptions with temporally variable plume altitudes, such as the 2010 Eyjafjallajökull, 2011
Nabro and 2014-15 Holuhraun eruptions.
3.7 Volcanic activity and eruption altitude information
31 ACCEPTED MANUSCRIPT
Information on eruption occurrence, timing and VEI is derived from the Volcanoes of the World
(VOTW) database curated by the Smithsonian Institution Global Volcanism Program (GVP; http://www.volcano.si.edu; Global Volcanism Program, 2013). The VOTW database is updated frequently (see http://www.volcano.si.edu/gvp_votw.cfm for latest updates) but for this analysis we have used VOTW v4.3.0, which includes eruptions through June 2014 (Global Volcanism
Program, 2013). It is important to note that VEIs provided in the VOTW database (particularly for more recent eruptions) are assigned based on the best available information at the time (e.g., plume altitude, estimated eruption volume) and hence are subject to some uncertainty.
Furthermore, a single VEI cannot capture any variability of explosive activity during long-lived eruptions; in such cases the assigned VEI corresponds to the largest event recorded in given time period.
The altitude of volcanic SO2 injection determines its climate impact, and hence our database includes estimates of SO2 altitude for all eruptions (Fig. 1). As discussed above, altitude estimates can be derived from several sources, with more frequent measurements of volcanic plume altitude available since 2002 (MIPAS), 2004 (MLS) and particularly sine June 2006
(CALIOP and CloudSat). CALIOP, launched in April 2006 on the Cloud-Aerosol Lidar and
Infrared Pathfinder SatelliteACCEPTED Observations (CALIPSO) MANUSCRIPT satellite (Winker et al., 2009), has been of tremendous value to the volcano remote sensing community with its ability to determine precisely the altitude of aerosols and clouds. CALIOP browse and expedited browse imagery products produced at NASA Langley Research Center (http://www- calipso.larc.nasa.gov/products/) are routinely checked during eruptions for evidence of plume detection. Lidar backscatter data in conjunction with depolarization ratios (indicating aerosol phase) and color ratios (indicating relative particle size) can be used to confidently distinguish
32 ACCEPTED MANUSCRIPT
volcanic plumes from meteorological clouds (e.g., Winker et al., 2012), especially when matched with coincident nadir satellite observations from the A-Train (e.g., OMI or AIRS SO2 data; Carn et al., 2009a, 2009b; Vernier et al., 2013). Although CALIOP detects aerosol particles and not
SO2 gas, it is often assumed that aerosols and SO2 will be approximately collocated (horizontally and vertically). Clarisse et al. (2014) found good agreement (within ~2 km in the vertical) between IASI SO2 altitude retrievals and CALIOP aerosol layer altitudes for the 2011 Nabro eruption cloud. However, collocation may not always be a valid assumption and some separation between gas and aerosol phases is expected due to different sedimentation rates. The poor spatial sampling of CALIOP along the A-Train orbit track often precludes detection of fresh volcanic eruption plumes (~1-3 hours old; where ash and hydrometeors may dominate; e.g., Rose et al.,
2000) until they expand in size in subsequent days, although the February 2014 eruption of Kelut
(Java, Indonesia) provided a rare CALIOP profile through an explosive eruption column only ~2 hours after emission (Kristiansen et al., 2015). The Cloud-Profiling Radar (CPR) on the
CloudSat satellite complements CALIOP with millimeter-wave (W-band; 3.2 mm wavelength) observations of larger cloud particles and precipitation, but is only sensitive to very fresh volcanic eruption clouds with high loadings of coarse ash and hydrometeors; to date very few confirmed CPR detectionsACCEPTED of volcanic clouds haveMANUSCRIPT been reported but there is evidence that a young plume (~4 hours old) from Redoubt (Alaska) was detected on March 26, 2009 (S.A. Carn, unpublished data). Unfortunately, no CloudSat data are available for the February 2014 Kelut eruption. More typically, CALIOP samples volcanic plumes several hours after eruption at the earliest, but for the larger eruptions CALIOP observations can continue for many days or weeks as SO2 gradually converts to sulfate aerosol and/or fine volcanic ash persists (e.g., Eckhardt et al., 2008; Shibata and Kouketsu, 2008; Carn et al., 2009a, 2009b; Spinei et al., 2010; Tulet and
33 ACCEPTED MANUSCRIPT
Villeneuve, 2011; Carn and Lopez, 2011; Vernier et al., 2011; Winker et al., 2012; Wang et al.,
2013; Fromm et al., 2014; Clarisse et al., 2014; Carn et al., 2015). CALIOP observations of volcanic clouds since 2006 have also revealed the complex, multi-layered structure of some plumes (Fig. 2; e.g., Carn et al., 2009a), which has implications for the altitude assumptions made in nadir satellite SO2 retrievals.
SO2 altitude retrievals from nadir UV or IR satellite observations (e.g., Yang et al.,
2009b, 2010; Nowlan et al., 2011; Rix et al., 2012; Carboni et al., 2012; Clarisse et al., 2014) offer the advantage of much better spatial sampling than CALIOP, although nadir-viewing satellite measurements are sensitive to a range of altitudes rather than a specific level; hence the retrieved SO2 altitude typically represents the peak of a larger vertical distribution. SO2 altitude retrievals require a minimum SO2 column amount but are now routinely performed using IASI data (Fig. 7; e.g., Clarisse et al., 2014), although to date UV SO2 altitude retrievals have only been applied to a small subset of eruptions (Yang et al., 2009b, 2010; Rix et al., 2012).
Prior to 2004-2006, and for many smaller eruptions since then which go undetected by
CALIOP and/or provide insufficient signal for SO2 altitude retrievals, the main sources of information on eruption altitude are Smithsonian GVP volcanic activity reports (which include pilot reports, satelliteACCEPTED observations, and ground MANUSCRIPT-based observations) and trajectory modeling.
4. Results and discussion
In presenting the results of our analysis, it is important to stress the ephemeral nature of any database pertaining to active volcanic processes. In essence, this work provides a snapshot of a dynamic inventory that is being continually revised as new eruptions occur or old events are reanalyzed (Carn, 2015c). We stress however, and demonstrate below, that the largest eruptions
34 ACCEPTED MANUSCRIPT
have been subject to greater scrutiny and are well-characterized using numerous satellite instruments; it is the smaller eruptions that are the source of most uncertainty. For larger eruptions, it may be possible to validate satellite SO2 measurements using the mass of sulfate aerosol produced; e.g., for the 1982 El Chichón eruption the TOMS estimate of SO2 loading is consistent with independent estimates of the mass of sulfate aerosol produced by SO2 oxidation
(Krueger et al., 2008). However, this is not possible for smaller eruptions. Although many eruptions are analyzed in close to real-time as they occur, providing preliminary estimates of SO2 loading, post-analysis of eruptions may continue many years or decades after the event as new retrieval algorithms are developed (e.g., Guo et al., 2004; Krueger et al., 2008). Table 2 lists many of the eruptions that have been subject to more detailed analysis (typically the larger events); this invariably results in more accurate estimates of SO2 loadings and uncertainties. We therefore acknowledge that significant uncertainty remains on SO2 loadings for some eruptions; estimates could potentially change by 50-100% in some cases though order of magnitude changes are considered extremely unlikely.
4.1 Eruptive degassing
ACCEPTED MANUSCRIPT
4.1.1 General trends in eruptive degassing
An overview of the entire eruptive volcanic SO2 inventory (Carn, 2015c) is given in Figure 1, with more detailed time-series of OMI (September 2004 – December 2014), AIRS (January 2003
– December 2014) and IASI (October 2007 – December 2014) SO2 data shown in Figures 4, 5, 6 and 7, and a map of average IASI SO2 columns depicting the geographical extent of detected volcanic SO2 plumes shown in Figure 8. Figure 9 shows a plot of cumulative SO2 emissions
35 ACCEPTED MANUSCRIPT
against time for different eruption styles. In total, the database currently contains SO2 measurements for ~700 eruptions or eruptive events, with a cumulative SO2 emission of ~101
Tg. Figures 1 and 9 display the main trends in eruptive SO2 emissions observed since 1978, namely the large perturbations associated with the 1982 El Chichón and 1991 Pinatubo eruptions, a significant decline after Pinatubo until the late 1990s, followed by a general increase throughout the 2000s, with peaks in 2005, 2008 and 2011, and reduced emissions in 2012-14.
Figure 9 shows possible evidence for an increase in the rate of SO2 production (primarily from explosive eruptions) beginning in 2004-2005, after a lull between the mid-1990s and early
2000s, although we interpret this cautiously since the change in slope corresponds to the onset of more sensitive SO2 measurements by OMI, and hence may simply reflect an increased rate of eruption detection (the mean detection rate increases from 8 eruptions/year in 1978-2004 to ~40 eruptions/year in 2004-2014). In contrast, the time-averaged flux of SO2 from effusive eruptions has been less variable over the past 3-4 decades, although a possible change in slope occurred in the early 2000s (Fig. 9), which may be partly related to an increased magma output rate at
Nyamuragira (DR Congo) in this timeframe (e.g., Wadge and Burt, 2011).
No major eruptions were reported during the 1995-96 TOMS data gap (Fig. 1), but we have assessed severalACCEPTED smaller events for SO2 emissionsMANUSCRIPT using HIRS/2 data. An effusive eruption of Fernandina (Galápagos Islands) in January-March 1995 was not detected by HIRS/2 (Head,
2005), indicating that SO2 emissions were confined to the lower troposphere. Some of the largest eruptions of the 1995-96 period occurred at Ruapehu (New Zealand), with significant SO2 degassing suggested by ground-based COSPEC measurements (Christenson et al., 2000) and eruption plume altitudes of up to ~9 km (Prata and Grant, 2001), which is close to the tropopause in the mid-latitude southern hemisphere winter. However, Prata et al. (2003) report
36 ACCEPTED MANUSCRIPT
relatively minor SO2 loadings of ~0.01-0.03 Tg for the June 17, 1996 eruptions. Analysis of
HIRS/2 data for other eruptions in 1995-96 has not revealed any significant UTLS SO2 emissions. We therefore conclude that the apparent decline in total volcanic SO2 emissions between 1994 and 1996 (Fig. 1) is genuine; an inference that is also supported by the lack of major SAOD perturbations in that period (Fig. 9).
The latitudinal distribution of eruptive SO2 emissions in 1978-2014 (Fig. 1) simply reflects the geographic distribution of active volcanoes, which has clusters at 0º-10ºS (including
Ecuador, the Galapagos islands, DR Congo, Indonesia and Papua New Guinea), 10º-20ºN
(including Mexico, central America, the West Indies, Afar, the Philippines, the Marianas, and
Hawaii), and 50º-60ºN (including Kamchatka, the northern Kuriles and the Aleutian islands), with a smaller peak at 30º-40ºN (including Sicily, Italy and Japan; Siebert et al., 2010). Regular eruptive activity in the equatorial region is evident, largely due to frequent effusive eruptions of
Nyamulagira (DR Congo; Carn and Bluth, 2003; Bluth and Carn, 2008) and the Galápagos
Islands volcanoes (Head, 2005; Carn, 2015). The spate of increased activity in southern Chile since 2008 (eruptions of Llaima, Chaitén, Cordon-Caulle, Copahue, and Calbuco) is also noteworthy in the otherwise less active extra-tropical southern hemisphere. The general dearth of southern hemisphereACCEPTED eruptions contrasts with theMANUSCRIPT high-latitude northern hemisphere, where there have been relatively frequent eruptions in the Kuriles, Kamchatka, Alaska and Iceland, particularly since 2008 (Fig. 1, 6, 8).
The more sensitive SO2 measurements by OMI since 2004 (Fig. 4) and IASI since 2007
(Fig. 7) provide more detailed insight into volcanic SO2 emissions in the last decade. As previously discussed, use of the maximum daily OSI to detect volcanic SO2 emissions in a continuous daily time-series (Fig. 4) avoids complications when SO2 from multiple sources at
37 ACCEPTED MANUSCRIPT
different altitudes is present (accurate SO2 loadings are shown in Figure 1). Spikes in the OSI correlate with known eruptions as indicated on Figure 4, with the OSI value roughly proportional to the maximum SO2 column amount and altitude on a given day, and the larger (or longer-lived) eruptions have distinctive OSI tails due to the persistence of SO2 in the atmosphere for days or weeks as it converts to sulfate aerosol. Comparison of the OSI with coincident maximum daily
SO2 and HCl mixing ratios measured by MLS (Fig. 4) clearly indicates the major UTLS injections (e.g., Manam in February 2005, Soufriere Hills in May 2006, Rabaul in October 2006,
Kasatochi in August 2008, Redoubt in March 2009, Sarychev Peak in June 2009 (see animation provided as Supplementary Material), Merapi in November 2010, Nabro in June 2011 and Kelut in February 2014; Table 2), although many smaller injections are also apparent upon closer inspection (see also Pumphrey et al., 2015). The significance of the observed HCl emissions will be discussed later (section 4.1.6). The advantage of SO2 measurements from multiple platforms is also demonstrated by the large eruption of Manam on November 24, 2004: this occurred during an OMI data gap but the corresponding UTLS SO2 and HCl injection was detected by
MLS (Fig. 4) and the SO2 cloud was measured by AIRS (Fig. 5), though it was not detected by
MIPAS (Höpfner et al., 2015).
Much of the backgroundACCEPTED variation in the MANUSCRIPT OSI time-series in periods devoid of large eruptions is due to smaller eruptions or very high passive SO2 emissions from volcanoes such as
Nyiragongo (DR Congo; 2004-2005; Carn, 2004), Ambrym (Vanuatu; 2005; e.g., Bani et al.,
2009, 2012), and Kilauea (Hawaii; strong summit degassing from March 2008), or extended periods of unrest at volcanoes such as Anatahan (CNMI; April – September 2005; see animation provided as Supplementary Material) and Kliuchevskoi (Kamchatka; May-July 2007). Section
4.2 discusses observed sources of passive SO2 degassing in more depth. It is also likely that some
38 ACCEPTED MANUSCRIPT
seasonally variable anthropogenic SO2 emissions are detected by the OSI: for example, the strong SO2 emissions from the Norilsk metal smelter (Siberia, Russia; e.g., Walter et al., 2012) produce a strong signal in the OMI data in the northern hemisphere spring (March-April) each year (Fig. 4), due to the presence of SO2 above highly-reflective snow-covered ground.
The AIRS and IASI measurements of UTLS SO2 emissions and altitude (Fig. 5, 6, 7) are broadly consistent with the OMI SO2 data, and also with MIPAS SO2 observations (Höpfner et al., 2015) with the 2008 Kasatochi, 2009 Sarychev Peak and 2011 Nabro eruptions dominating the time-series in terms of SO2 loading (Table 2). IASI SO2 altitude retrievals for these eruptions indicate injection of SO2 to a range of altitudes in the UTLS (Fig. 7), which is consistent with
MLS data (Fig. 4). These vertical profiles contrast with the more localized UTLS SO2 injections associated with the tropical eruptions of Soufriere Hills in May 2006, Rabaul in October 2006,
Merapi in November 2010 and Kelut in February 2014 (Fig. 4, 7); this may be a consequence of variable eruption magnitude and style, or of diverse atmospheric dynamics in the tropics and mid-latitudes. An advantage of IASI and AIRS is that they provide robust measurements of high- latitude winter eruptions, such as the December 2010 eruption of Kizimen (Kamchatka; which was incorrectly ascribed to Shiveluch by Höpfner et al. (2015)), the November 2012 eruption of
Tolbachik (e.g., TellingACCEPTED et al., 2015), and the 2014MANUSCRIPT eruption of Holuhraun (Bardabunga, Iceland;
Fig. 5, 7). Long-range dispersal of SO2 emitted by the June 2011 eruption of Cordon Caulle
(Chile), which occurred in the southern hemisphere winter, was also tracked more effectively by the IR sensors (Fig. 6, 8). Average SO2 columns detected by IASI clearly show the imbalance between northern and southern hemisphere SO2 emissions since 2007, and long-range tracking of
SO2 from several eruptions made possible by sensitivity to low SO2 column amounts (Fig. 8).
During a two-month period in May-June 2011, large eruptions occurred in Iceland (Grimsvötn),
39 ACCEPTED MANUSCRIPT
Chile (Cordon Caulle) and Eritrea (Nabro), resulting in UTLS SO2 emissions spread over an unusually large latitudinal range (Fig. 6).
Comparisons between satellite SO2 measurements in the overlap periods between missions provide confidence in the data for larger eruptions and in the multi-decadal continuity of the volcanic SO2 database (e.g., Table 2). TOMS, MODIS, AIRS and MIPAS measured very similar SO2 amounts (~0.1-0.12 Tg) in the July 2003 Soufriere Hills eruption cloud, which reached 16 km altitude (Table 2; Prata and Bernardo, 2007; Carn and Prata, 2010; Höpfner et al., 2015). The UTLS SO2 injection from Anatahan on April 6, 2005 was observed by TOMS,
OMI, AIRS and MIPAS, with each sensor measuring ~70-80 kt SO2 (Table 2; Höpfner et al.,
2015), with similar agreement observed for most other UTLS SO2 emissions. Other eruptions, such as the October 2005 effusive eruption of Sierra Negra (Galapagos Islands) show discrepancies between satellite measurements (Table 2) but often there are good reasons for this; in this case a significant fraction of the emitted SO2 was lower tropospheric, preventing detection by TOMS, which also had some data gaps. Variable spatial coverage of volcanic clouds can also produce differences (e.g., Prata and Bernardo, 2007), hence in general the largest SO2 mass measured by any sensor is assumed to provide the best estimate.
The August 2008ACCEPTED eruption of Kasatoch MANUSCRIPTi, one of the largest explosive eruptions since
Pinatubo, is a good example of an event that has been subject to detailed analysis with encouraging convergence of total SO2 loading estimates from multiple satellite instruments and analysis techniques (Table 2). Contemporary eruption reports suggested ash emissions to at least
11-14 km altitude during the Kasatochi eruption on August 7, 2008 (Global Volcanism Program,
2008). Multiple satellite instruments observed the emitted SO2 cloud. Kristiansen et al. (2010) used SO2 measurements from OMI, GOME-2 and AIRS in an inversion scheme to derive a total
40 ACCEPTED MANUSCRIPT
SO2 discharge of 1.7 Tg, and a co-eruptive vertical SO2 profile extending from 7-20 km altitude with a major peak at 12 km. They also estimated that ~1 Tg SO2 (i.e., ~60% of the total SO2 discharge) was injected into the local stratosphere (>10 km altitude), although this is dependent on the tropopause definition (Kristiansen et al., 2010; Clarisse et al., 2014; similarly, Eckhardt et al. (2008) estimated ~10-60% of SO2 emitted by the October 2007 eruption of Jebel al-Tair
(Yemen) was injected into the stratosphere, depending on tropopause definition). Kristiansen et al. (2010)‘s SO2 loading estimate for Kasatochi could be regarded as robust since the inversion technique includes multiple UV and IR satellite datasets. Individual satellite datasets yield similar results: Karagulian et al. (2010) report 1.7 Tg SO2 based on IASI measurements using both the 7.3 µm and 4 µm SO2 bands. A subsequent, improved analysis of IASI data by Clarisse et al. (2012) found a maximum of 1.6 Tg SO2 measured on Aug 11, 2008; in agreement with the
1.6 Tg SO2 measured by GOME-2 (Nowlan et al., 2011). Prata et al. (2010) retrieved a total of
~1.2 Tg SO2 using AIRS data, with the lower SO2 amount detected perhaps due to lower sensitivity to low SO2 columns and incomplete cloud coverage by AIRS. Krotkov et al. (2010) report a slightly higher value of ~2.2 Tg SO2, injected to 10-12 km altitude, based on OMI measurements and trajectory modeling; in this case the estimated SO2 loading was derived from extrapolation of the SOACCEPTED2 decay trend back to the MANUSCRIPT eruption time. This procedure is subject to some uncertainty as the prevailing SO2 removal processes may not be the same in the early and aged volcanic plume, plus the observed e-folding time for SO2 removal was relatively short (~9 days;
Krotkov et al., 2010), possibly due to proximity to the tropopause, and this impacts the extrapolated SO2 mass estimate. Note that the largest SO2 mass measured by OMI in a single, coherent cloud was 1.8 Tg on August 10 (Krotkov et al., 2010). MODIS SO2 retrievals for
Kasatochi, including 8.6 µm data corrected for ash interference (Corradini et al., 2010), are
41 ACCEPTED MANUSCRIPT
broadly consistent with the other estimates within stated uncertainties (Table 2). Thus, when all
UV and IR satellite measurements are considered we find general convergence on a total SO2 emission of ~2 Tg for the 2008 Kasatochi eruption, a value that can be considered very robust.
4.1.2 Eruptive SO2 flux
Knowledge of the time-averaged flux of volcanic SO2 into the UTLS is crucial for interpretation of trends in SAOD and assessment of the impact of volcanic activity on climate. Using TOMS
SO2 data for 1978-1993, Bluth et al. (1993) estimated an annual volcanic SO2 flux from explosive volcanism of ~4 Tg, clearly influenced by the large El Chichón and Pinatubo eruptions in that era. Pyle et al. (1996) calculated a medium-term (~100 year) flux of ~1 Tg/yr volcanic
SO2 to the stratosphere (0.3-3 Tg/yr range), and a short-term (10-20 year) flux due to small eruptions (1010-1012 kg or VEI 3-5) of ~0.4 Tg/yr. Here, we provide updated estimates of the volcanic SO2 flux to the UTLS based on recent satellite measurements, with improved data for smaller eruptions.
Table 3 reports the total annual SO2 emissions measured by satellites from 1978 to 2014 for all detected eruptions and for explosive eruptions only (i.e., excluding effusive eruptions; also plotted in Fig. 1). TheACCEPTED latter have greater potential MANUSCRIPT to reach the UTLS and impact climate. Our distinction between explosive and effusive eruptions in the SO2 emissions database (Carn,
2015c) is based on the dominant style of activity; we define effusive eruptions as those dominated by effusion of basaltic lava flows (typically in intraplate settings). Eruptions involving extrusion of silicic lava domes (e.g., Soufriere Hills, Montserrat) are categorized as
‗explosive‘ since it is the discrete, explosive emissions from these eruptions that are most effectively measured from space. To estimate total SO2 emissions from effusive eruptions we
42 ACCEPTED MANUSCRIPT
have used the procedure described by Bluth and Carn (2008), which accounts for rapid, daily removal of SO2 in lower tropospheric plumes.
A notable feature of the volcanic SO2 data is significant inter-annual variability, with total explosive SO2 emissions varying from 10 kt (in 1987) to ~24 Tg (in 1991; Table 3). On average, ~50% of annual measured SO2 emissions have been sourced from explosive eruptions
(Table 3), indicating the important contribution of sulfur-rich effusive basaltic eruptions to the global flux. The mean annual SO2 fluxes from all detected eruptions and explosive eruptions are
2.8 Tg (range = 0.04-26 Tg) and 1.6 Tg (range = 0.01-24 Tg), respectively; the latter can be considered an estimate of the SO2 flux to the UTLS, and if we assume that ~60% of this SO2 enters the stratosphere (based on detailed analyses of the 2008 Kasatochi and 2007 Jebel al-Tair eruptions; Eckhardt et al., 2008; Kristiansen et al., 2010) then the stratospheric flux is ~1 Tg/yr
3 SO2, similar to earlier estimates of medium-term and longer-term (~10 years) fluxes (Pyle et al.,
1996). We can conclude that the flux of volcanic SO2 to the stratosphere in the past 3-4 decades has not been abnormally high or low. The relationship between SO2 emission and eruption magnitude is explored in more detail in section 4.1.3.
The exact proportion of emitted SO2 that enters the stratosphere during an explosive eruption will vary withACCEPTED eruption magnitude, location MANUSCRIPT and style. Although we attempt to separate volcanic SO2 injected into the UTLS from that injected into the stratosphere sensu stricto, volcanic SO2 injected into the upper troposphere below the tropopause also has the potential to impact the stratosphere. Effusive eruptions are not generally regarded as significant sources of stratospheric pollution, notwithstanding large-volume fissure eruptions with high eruption rates
(e.g., Stothers et al., 1986), yet they frequently emit prodigious quantities of SO2 into the upper troposphere (e.g., Krueger et al., 1996; Carn and Bluth, 2003; Head, 2005; Bluth and Carn,
43 ACCEPTED MANUSCRIPT
2008; Hughes et al., 2012; Höpfner et al., 2015) where it has the potential to cross the tropopause. MIPAS detected SO2 in the UTLS (at altitudes of up to 18-22 km) after several eruptions of Nyamuragira (in 2002, 2006 and 2010) and the 2009 Fernandina eruption
(Galápagos Islands; Höpfner et al., 2015), suggesting that such eruptions should not be neglected as a possible source of stratospheric SO2. The June 12, 2011 eruption of Nabro, which was initially (and, it transpired, incorrectly) reported as tropospheric but produced a significant
SAOD anomaly (e.g., Sawamura et al., 2012; Uchino et al., 2012; Bourassa et al., 2012;
Mateshvili et al., 2013), ignited a debate over the relative roles of direct injection and cross- tropopause transport by other processes such as deep convection or isentropic ascent in lofting volcanic SO2 into the stratosphere (Bourassa et al., 2012; Fromm et al., 2013; Vernier et al.,
2013; Bourassa et al., 2013; Fairlie et al., 2014; Fromm et al., 2014; Penning de Vries et al.,
2014; Clarisse et al., 2014). In addition to the evidence presented in these works, which indicate that both early direct injection and subsequent slow ascent of SO2 and sulfate aerosol contributed to Nabro‘s stratospheric impact, daily MLS SO2 retrievals (Fig. 10) also show SO2 close to or above the tropopause (at 68 hPa) as early as June 13-14 plus a slower ascent of SO2 (up to ~46 hPa or 19-20 km altitude) in the subsequent two weeks. Interestingly, MLS data do not show the same evidence for substantialACCEPTED stratospheric HCl MANUSCRIPT input (Fig. 10), likely due to more efficient HCl removal processes in the troposphere. Overall, the Nabro eruption demonstrates that volcanic eruptions can be complex, multifaceted events and require analysis on a case-by-case basis to unravel their stratospheric impact. The Nabro eruption displayed characteristics of both explosive and effusive eruptions in that it initially produced an energetic eruption column (e.g.,
Fee et al., 2013) that penetrated the UTLS, yet it did not produce a substantial ash cloud and emitted a long lava flow, with tropospheric SO2 emissions continuing for several weeks after the
44 ACCEPTED MANUSCRIPT
initial paroxysm (e.g., Theys et al., 2013; Goitom et al., 2015). It can thus be considered a relatively rare, hybrid event sharing features of two ‗end-member‘ eruptions: a sulfur-rich, effusive eruption injecting SO2 predominantly into the troposphere (e.g., similar to Nyamuragira) and an explosive eruption with direct stratospheric injection (e.g., the January 2005 Manam eruption; Fig. 4).
4.1.3 VEI-SO2 loading relationships
Analysis of the relationship between SO2 production and eruption magnitude (VEI) can inform studies of the atmospheric and climate impact of eruptions prior to the satellite measurement era.
In comparing measured SO2 emissions with reported eruption VEIs, we extend earlier work based on TOMS SO2 data by Schnetzler et al. (1997) with the addition of many more observations of smaller eruptions due to improved satellite data quality. We also attempt to estimate how much volcanic SO2 could elude detection from space based on VEI-SO2 mass relationships. We note that more precise (i.e., continuous) measures of eruption size, such as magnitude and intensity, are more useful for such analyses than the more limited cardinal (or discrete) VEI scale, since each VEI category encompasses a significant range of eruption size
(e.g., Pyle, 1995, 2015).ACCEPTED However, magnitudes MANUSCRIPT and intensities are only available for a relatively small number of eruptions, and here we use the Smithsonian GVP database of VEIs from VOTW
4.3.0. Some small modifications to the VOTW database were necessary to account for long- duration eruptions assigned a single VEI; hence continuous activity at Anatahan in 2005, included as a single VEI 3 eruption in the VOTW catalog, was subdivided into discrete VEI 2 and 3 events based on reported plume altitudes.
45 ACCEPTED MANUSCRIPT
As pointed out by Pyle (2015), there are few metrics other than VEI, magnitude and intensity that are routinely used to describe eruption size. However, we contend that SO2 measurements are the most accurate way of quantifying the impact of an eruption on climate
(and atmospheric composition generally), which is ultimately among the main goals of any eruption magnitude scale, given that volcanic ash has only minor climate impacts (e.g., Robock,
2000). Since SO2 emissions are not clearly linked to other magnitude scales (which are more relevant to impacts determined by emissions of solid ejecta such as volcanic ash and lava, e.g., on aviation or the environment), except in the broadest sense (e.g., Schnetzler et al., 1997; Blake,
2003; Scaillet et al., 2003), there may be a need for multiple measures of eruption magnitude and impact. Intensity (which determines eruption column altitude) may be a more relevant metric than magnitude or VEI for assessment of climate impacts.
Tables 4 and 5 summarize the measured SO2 emissions per VEI category for explosive eruptions and all detected eruptions, respectively, and the data for individual eruptions are plotted in Figure 11. Figure 11 indicates that, when effusive eruptions (as defined in section
4.1.2) are excluded, there is a broad correlation between VEI and SO2 emission; i.e., there is roughly an order of magnitude increase in mean SO2 discharge with increasing VEI (Table 4). A similar correlation wasACCEPTED found by Schnetzler et MANUSCRIPTal. (1997) using a smaller dataset. Based on the mean SO2 loading for each VEI (Table 4) we find the following relationship:
log10SO 2 , Tg 0.71 VEI 3.15 1
Cross-checking with the VOTW database shows that SO2 emissions by all reported eruptions with VEI≥4 have been quantified (consisting of 26, 4, and 1 eruptions of VEI 4, 5, and 6,
46 ACCEPTED MANUSCRIPT
respectively, reported in the VOTW database), confirming that satellite measurements provide excellent constraints on SO2 emissions by the largest eruptions. Note that some eruptions are assigned high VEIs due to large cumulative erupted volumes over long periods (years), but may not feature any discrete large eruptive events; e.g., activity at Kliuchevskoi (Kamchatka) in
1986-1990 (where SO2 was occasionally detected by TOMS), or long-term dome-building eruptions at Shiveluch (Kamchatka) in 1999-2014, during which time a few relatively small SO2 emissions have been detected. In such cases we include these eruptions as ‗detected‘ if any SO2 emissions were measured at any time during the reported eruption period, although the total eruptive SO2 emissions may be poorly constrained. Emissions from such continuous eruptions are more akin to passive volcanic degassing, with similar challenges for detection from space.
For eruption magnitudes below VEI 4, the SO2 emission record becomes increasingly incomplete. Based on reported VEIs, around 50% of VEI 3 eruptions (n = 162) have been detected by satellite SO2 measurements (Table 4, 5). The VEI 3 category likely includes some phreatic or phreatomagmatic eruptions that may generate substantial eruption columns (>3 km altitude required for VEI 3) but which would not emit large quantities of SO2. Some VEI 3 events also correspond to extended periods of volcanic unrest, often with continuous SO2 emissions (e.g., AmbrymACCEPTED, Vanuatu). Other VEI MANUSCRIPT 3 eruptions only appear to produce ash, with no detectable SO2; e.g., many eruptions of Bezymianny (Kamchatka) produce little measurable SO2, but based on Smithsonian GVP reports it appears that the larger explosive eruptions are detected.
However, Bezymianny clearly does not appear to be a major source of SO2 emissions based on detected events. We also stress that, for the lower magnitude eruptions, not all satellite data has been analyzed in detail due to the increasingly large number of events. However, it is likely that any significant SO2 emissions would have been noticed and recorded.
47 ACCEPTED MANUSCRIPT
At magnitudes of VEI≤2, it is clear that the majority of eruptions (87% or more) are not detected (Table 4, 5), most likely due to insufficient SO2 release. However, it is also certain that not all the >1000 reported VEI 0-2 eruptions have been checked for SO2 emissions in satellite data. By definition, non-explosive, effusive eruptions (e.g., in Hawaii, DR Congo, the Galápagos
Islands and Iceland) are generally assigned a VEI of 0-2, but these are all detected due to their typically SO2-rich emissions (e.g., Carn and Bluth, 2003; Sharma et al., 2004; Head, 2005; Bluth and Carn, 2008; Carn, 2015b), which obfuscate any relationship between SO2 emission and VEI
(e.g., Fig. 11). In fact, the only VEI 0 eruptions with confirmed SO2 emissions are effusive eruptions (Table 4, 5; Fig. 11). Although the undetected, explosive VEI 0-2 eruptions are unlikely to be significant from a climate perspective (due to low column heights and/or low SO2 amounts), quantifying their emissions may be important for the global sulfur budget and sulfur cycle, and for the budgets of other, longer-lived volcanic gases such as CO2, but would require more extensive ground-based measurements of this class of eruption. Indeed, at the lower end of the VEI scale the distinction between continuous, passive degassing and frequent small eruptions becomes tenuous.
Despite the poor sampling of low magnitude eruptions, a notable feature of the SO2 data is the high variabilityACCEPTED of SO2 emissions from eruptionsMANUSCRIPT of intermediate size (VEI 2-4), with standard deviations similar to or exceeding the mean SO2 loading for each VEI (Table 4, 5; Fig.
11). The range of SO2 emissions for intermediate VEIs is larger than reported by Schnetzler et al.
(1997) due to improved data for smaller eruptions; thus the increased variability may reflect sampling of eruptions with a larger range of magmatic sulfur contents. Since all VEI 4 eruptions since 1978 have been detected, we have good constraints on the variability of SO2 discharge from this magnitude class, which covers ~2-3 orders of magnitude. The greatest variability is
48 ACCEPTED MANUSCRIPT
found for VEI 3 eruptions (Table 4, 5), which range over ~3 orders of magnitude (Fig. 10). The actual variability may be even greater, since only 50% of VEI 3 eruptions have been detected
(likely the more SO2-rich ones). This extreme variability makes it challenging to estimate SO2 emissions from past eruptions based on VEI alone, but this analysis at least provides some potential error bars on such predictions. One likely cause of the observed variability is that some
VEIs may be inaccurate, which relates to our earlier point that SO2 emissions offer a useful alternative indication of eruption magnitude. For example, both the 2011 Nabro and 2008
Kasatochi eruptions were assigned a VEI of 4, but their total SO2 emissions (~2-4 Tg) are the highest in this VEI class and are similar to or greater than the mean SO2 loading for VEI 5 eruptions. Indeed, for Nabro, Goitom et al. (2015) estimate a magnitude of 5.1 based on total observed SO2 emissions (Theys et al., 2013) and magmatic sulfur content. Similarly, the magnitude of the Kasatochi eruption has been estimated as 4.5-4.7 (Pyle, 2015), although we note that the 2008 Chaitén eruption (which produced <20 kt of detectable SO2) was also a magnitude 4.7-5 event. Nevertheless, refinement of eruption magnitudes would likely improve the correlation between SO2 emission and VEI shown in Figure 11, and the SO2 measurements provide perhaps the most robust basis for adjusting VEIs for recent eruptions. Finally, we note that Scaillet et al. (2003)ACCEPTED also found an increased MANUSCRIPT spread of estimated SO2 yield with decreasing eruption magnitude based on petrological data, which they attributed to the variability in bulk sulfur contents of magmas. Thus the variability in observed SO2 loadings for smaller eruptions
(Fig. 10) may represent sampling of magmas with different bulk sulfur content.
To assess the SO2 emissions from low-magnitude explosive eruptions that elude detection from space, we can use the mean SO2 loading for detected eruptions, the VEI-SO2 regression
(equation [1]), or an estimate of the sensor detection limit. For example, the VOTW 4.3.0 catalog
49 ACCEPTED MANUSCRIPT
includes 507 eruptions with VEI 1, of which only 4% have detected SO2 emissions (Table 4), i.e., 487 eruptions are unaccounted for. Using the mean SO2 emission for detected VEI 1 eruptions (excluding effusive eruptions) of 0.006 Tg SO2 yields an estimated SO2 emission of
487×0.006=2.9 Tg SO2 from undetected eruptions (Table 4). Using equation (1) yields an estimated SO2 loading of 0.004 Tg per eruption and a total undetected amount of 1.9 Tg SO2. We consider these upper limits, since it is likely that the larger SO2 emissions in each VEI category would be detected, and hence the mean SO2 mass may be biased high, and some eruptions are known to have produced only ash emissions and/or no measurable SO2 (e.g., phreatic or phreatomagmatic eruptions, submarine eruptions). Using the same logic for all VEI categories with missing data (VEI 0-3) we calculate the estimates of undetected SO2 emissions and equivalent annual mean fluxes for the ~36 years of satellite measurements listed in Table 4.
Hence, the total estimated unmeasured SO2 discharge for VEI 0-3 eruptions is ~0.48 Tg/yr SO2, or ~17 Tg SO2 in total (Table 4), i.e., equivalent to a single Pinatubo-scale event, or ~20% of the
~100 Tg total measured SO2 emissions. Most of the undetected SO2 is sourced from VEI 2 eruptions (~50%; Table 4). Again, we consider these estimates to be upper bounds for the reasons mentioned above. The undetected SO2 emissions are commensurate with one VEI 4 eruption per year, or ACCEPTEDa single VEI 6 eruption, butMANUSCRIPT although this appears significant we stress that the unquantified SO2 is sourced from low-magnitude eruptions and hence is unlikely to significantly impact the stratosphere or climate. However, the emissions will impact the tropospheric SO2 budget and the budgets of other volcanic gases (e.g., CO2). Due to variable eruption rates, the unmeasured fraction may be more significant in some years than others, but we find an overall uncertainty of ~20% on total eruptive SO2 emissions to be very acceptable. A further consideration in calculating the mean SO2 emission for each VEI category is whether the
50 ACCEPTED MANUSCRIPT
probability density function (PDF) of SO2 loading for measured eruptions is Gaussian or lognormal, since this affects calculation of the mean. Although data are limited, the PDF appears approximately Gaussian for some VEIs (VEI 4), but may be more lognormal for lower VEIs.
Further statistical analysis of the database is encouraged.
In summary, we conclude that the flux to the stratosphere (or UTLS) from VEI 4+ eruptions is very well constrained (Table 4, 5), but significant uncertainty remains on the flux from smaller eruptions (as for passive volcanic degassing). However, more eruptions of low VEI eluded detection during the TOMS missions (1978-2004) than since 2004 with the onset of more sensitive OMI and IASI SO2 measurements, so the uncertainty on SO2 emissions from small eruptions is lower for the last decade.
4.1.4 Sources of eruptive volcanic SO2 emissions
In total, around 110 volcanoes are responsible for the eruptive SO2 emissions detected from space in the current database (Fig. 12). With the exception of Antarctica and the South Sandwich
Islands, at least one eruption has been detected in each of the 19 volcanic regions defined by the
Smithsonian GVP (Siebert et al., 2010; Fig. 12). A further ~30 volcanoes have produced passive
(non-eruptive) SO2 emissionsACCEPTED detected in satellite MANUSCRIPT data (see later discussion), resulting in a total of ~140 global volcanoes with detectable SO2 emissions since 1978. This is only ~9% of the
1545 volcanoes with known or inferred Holocene eruptions, or ~17% of the 858 volcanoes with dated Holocene eruptions, listed in the Smithsonian GVP catalog (Siebert et al., 2010). Table 6 lists the sources of detected eruptive SO2 emissions over 10 kt between 1978 and 2014 in decreasing order of total measured SO2 production, along with the number of eruptions generating the observed SO2 emissions, and a qualitative assessment of degassing style. Only
51 ACCEPTED MANUSCRIPT
eruptive SO2 emissions are reported in Table 6, i.e., passive degassing from volcanoes with variable degassing styles such as Etna and Kilauea is excluded. The global distribution of eruptive SO2 sources is shown in Figure 12.
The sources in Table 6 highlight the significant contribution of effusive volcanic eruptions to the eruptive volcanic SO2 flux (e.g., Nyamuragira, Sierra Negra, Holuhraun,
Fernandina, Cerro Azul, Mauna Loa, Wolf), though the SO2 emitted by these events is mostly tropospheric and less significant for climate. However, effusive eruptions have the potential to inject SO2 at or above tropopause altitudes, particularly at higher latitudes (e.g., Stothers et al.,
1986; Krueger et al., 1996), and have been linked to SAOD perturbations in the past (e.g., a
SAOD disturbance in 1979 was partly attributed to the 1979 Sierra Negra eruption; Fig. 13;
Rosen and Hofmann, 1980). Hence, given the frequency of these eruptions (e.g., Fig. 1, 9), they constitute an important source of SO2 in the UTLS. Otherwise, the dominant eruptive SO2 emissions (>2 Tg SO2) have been produced by the first recorded eruptions of previously dormant volcanoes (mostly in subduction zones), typically after centuries or more of quiescence (e.g.,
1991 Pinatubo: previous dated eruption in 1450; 1982 El Chichón: previous known eruption in
1360 or 1850; 2011 Nabro: first known eruption; 2008 Kasatochi: previous confirmed minor eruption in 1760; 2003ACCEPTED Anatahan: first documented MANUSCRIPT eruption; Siebert et al., 2010). However, there are exceptions: the large 1991 eruption of Cerro Hudson followed relatively recent, albeit smaller, eruptions in 1891 and 1971 (Siebert et al., 2010), although in this case the large SO2 emission in 1991 may have been supplied by a large basaltic dyke intrusion that triggered the eruption (e.g., Kratzmann et al., 2009). One possible explanation for the broad correlation between inter-eruptive quiescence and SO2 emission is that it reflects the accumulation of a vapor phase in long-lived magma chambers during repose (e.g., Wallace and Gerlach, 1994;
52 ACCEPTED MANUSCRIPT
Scaillet et al., 2003; Wallace, 2005). Although many of the dominant sources have limited data on inter-eruptive degassing (no significant SO2 emissions have been detected from space, but weak, fumarolic emissions would not be measured), they appear to release the vast majority of their SO2 during eruptions. Whether this SO2 originates from a stored vapor phase or syn- eruptive melt degassing is an open question that requires more extensive petrological studies.
However, the behavior of other volatile species such as CO2 may be different (e.g., there could be substantial inter-eruptive outgassing of CO2 exsolved at depth), but would require extensive ground-based CO2 flux measurements to resolve.
After the dominant explosive SO2 sources, which tend to produce single, large injections of SO2 into the UTLS, there are a number of volcanoes that produce smaller but more frequent emissions. The latter include Rabaul, Manam, Ulawun, Soputan, Kelut, Soufriere Hills, Sarychev
Peak and Hekla (Table 6), which generally produce <1 Tg SO2 per eruption but erupt relatively frequently (e.g., Tupper et al., 2004; Tupper et al., 2007; Carn and Prata, 2010; Kushendratno et al., 2012) and in some cases may be persistently active between larger eruptions (e.g., Rabaul,
Manam). Some volcanoes, such as Manam and Soputan, are prolific SO2 emitters with 100% eruption detection rates where explosive degassing appears to dominate the total SO2 flux, whereas passive SO2ACCEPTED emissions appear more significant MANUSCRIPT at others (e.g., Tungurahua (Ecuador) and Nevado del Ruiz (Colombia)). Soufriere Hills volcano (SHV) is unusual in that the SO2 emissions listed in Table 6 are all associated with lava dome collapses or explosive events during a single, long-lived dome-forming eruption (Carn and Prata, 2010); these events trigger the release of a deep, SO2-rich vapor phase (e.g., Christopher et al., 2010) and are not supplied by syn-eruptive degassing. Although these are not explosive eruptions in the conventional sense, they are still capable of injecting SO2 into the UTLS and impacting SAOD (e.g., the July 2003,
53 ACCEPTED MANUSCRIPT
May 2006 and February 2010 events at SHV; Fig. 4, 7, 13; Prata et al., 2007; Carn and Prata,
2010).
Smaller eruptive SO2 emissions may be a consequence of shorter inter-eruptive repose times, precluding accumulation of a magmatic gas phase, and/or persistent degassing between eruptions that may bleed off accumulating gases (Table 6). In general, volcanoes exhibiting higher levels of passive degassing appear less prone to large explosive eruptions, although even
Etna (one of the strongest sources of passive volcanic SO2 emissions; section 4.2) has a significant eruptive SO2 flux (Table 6). Emissions from frequent, moderate-sized eruptions dominate the short-term eruptive volcanic SO2 flux and in the tropics they have a higher climate impact than larger eruptions on a time-averaged basis due to their higher frequency (e.g., Miles et al., 2004). In the tropics, small eruptions can have disproportionate impacts on the UTLS due to vertical transport of SO2 assisted by deep convection (e.g., Tupper et al., 2009). Several of these modest tropical eruptions (e.g., 2002 Ruang, 2005 Manam, 2006 Soufriere Hills, 2006 Rabaul) have been responsible for small SAOD perturbations since 2000 (e.g., Fig. 9, 13), maintaining the stratospheric aerosol layer above background (non-volcanic) levels and likely playing a role in the global warming ‗hiatus‘ observed since 1998 (e.g., Solomon et al., 2011; Vernier et al.,
2011; Neely et al., 2013;ACCEPTED Ridley et al., 2014; Santer MANUSCRIPT et al., 2014; Carn et al., 2015).
A detailed examination of recent trends in SAOD (Fig. 9, 13) confirms the dominant influence of volcanic eruptions, but also shows that no major SO2 emissions have eluded detection by satellite instruments. All major perturbations to SAOD since 1979 can be attributed to a known volcanic eruption, and the more comprehensive record of volcanic SO2 emissions presented here also elucidates some of the more subtle SAOD variations apparent in Figs. 9 and
13. For example, the September 1994 eruption of Rabaul appears responsible for a small increase
54 ACCEPTED MANUSCRIPT
in SAOD above the post-Pinatubo background (Fig. 9, 13), and this and other eruptions (e.g.,
Manam in December 1996) may have prolonged the decay of the Pinatubo aerosol to background levels. It has been shown that the overall increase in SAOD observed since 2002 (Fig. 9, 13) is not due to increased anthropogenic SO2 emissions, as had previously been suggested (Hofmann et al., 2009), but to moderate-sized volcanic eruptions (Neely et al., 2013). The frequency of
UTLS SO2 injections since 2002 has been sufficient to prevent the return of tropical SAOD to background conditions, although a slight decline in SAOD is apparent from 2012-2014 (Fig. 9,
13) due to reduced levels of volcanic activity (at the time of writing, more recent SAOD data were not available to gauge the atmospheric impact of the large eruptions of Kelut in February
2014 and Calbuco in April 2015). We note, however, that there are still discrepancies between
SAOD observations and climate model simulations of stratospheric aerosol, even when the latter include forcing by the larger eruptions since 2002 (Neely et al., 2013). This appears to be due to a combination of incomplete records of volcanic SO2 forcing used to initialize model simulations, an over-emphasis on VEI as an indication of potential climate impact, and incorrect date assignments for some eruptions. For example, both Neely et al. (2013) and Santer et al.
(2014) list a VEI 3 eruption of Anatahan on April 12, 2004. Although this is the date given for the onset of the 2004ACCEPTED-2005 Anatahan eruption MANUSCRIPTby the Smithsonian GVP, it actually represents the beginning of an extended phase of activity, which peaked in April-August 2005 with several VEI
3 eruptive events and UTLS SO2 injections (Fig. 4, 5). Similarly, Vernier et al. (2011) ascribe minor stratospheric impact to an eruption of Shiveluch in May 2001 (based on its VEI of 4), but this produced no detectable SO2 emissions and we therefore contend that a smaller (VEI 3) eruption of Ulawun on April 30, 2001 (which injected at least ~30 kt SO2 into the tropical
UTLS) was more significant. Thus in these cases the satellite measurements provide the key
55 ACCEPTED MANUSCRIPT
information on the timing and location of SO2 emissions required for accurate climate model simulations. We expect that use of our comprehensive volcanic SO2 emissions database will likely resolve some of the remaining discrepancies between model simulations and observations, and also assist efforts to identify periods when the stratosphere was truly free of any volcanic influence.
4.1.5 Lifetimes of SO2 in the UTLS
The lifetime of SO2 in UTLS volcanic plumes (typically represented by an e-folding time; i.e.,
-1 the time for SO2 mass reduction by a factor of e ) is one of the key factors determining the climate impact of an eruption (along with SO2 amount, injection altitude and latitude; Robock,
2000). Observations of SO2 lifetime can also be used to test climate model simulations of volcanic plumes and improve the chemical schemes embedded in such models. With growing numbers of volcanic eruptions observed from space, we have increasingly accurate constraints on SO2 lifetimes in volcanic plumes at a range of altitudes and latitudes, and in all seasons.
Figure 14 shows a representative set of e-folding times for SO2 removal from volcanic plumes, based on IASI measurements for recent eruptions (since 2007) and values from the literature for older events and lowerACCEPTED tropospheric SO2 plumes MANUSCRIPT. Note that the e-folding times shown in Figure
14 characterize SO2 removal (e.g., via gas- or aqueous-phase oxidation, cloud processing or wet/dry deposition) and not sulfate aerosol removal from the UTLS, which also influences climate impact.
The data in Figure 14 confirm the strong correlation between SO2 altitude and lifetime, with increasing variability apparent for lower tropospheric plumes. However, it is not absolute altitude but rather altitude relative to the local tropopause that is the main determinant of e-
56 ACCEPTED MANUSCRIPT
folding time, as evinced by the high-latitude eruptions of Sarychev Peak, Kasatochi, Cordon
Caulle and Grimsvötn, amongst others (Fig. 14). These eruptions exhibit comparatively large e- folding times relative to the main trend defined by tropical eruptions, which is presumably due to the lower tropopause altitude at mid- to high-latitudes. Lower solar insolation at higher latitudes may also extend SO2 lifetimes by reducing the rate of some photochemical reactions (e.g., the production of OH-). The 1991 Cerro Hudson eruption, which is also mid-latitude, is somewhat of an anomaly (Fig. 14), but a shorter than expected SO2 lifetime in this case may be due to the timing of the eruption in the late southern hemisphere winter. With the exception of Cordon
Caulle (which also has a slightly lower e-folding time; Fig. 14) the more recent high latitude eruptions all occurred in the spring or summer. There is thus evidence for both altitude and seasonal impacts on the SO2 lifetime in volcanic plumes, which may be further elucidated as more eruptions are analyzed. The present dataset is too limited (and recent eruptions too small) to draw any conclusions regarding the effects of SO2 loading alone on SO2 lifetime, for example, if excessive stratospheric SO2 loading could sufficiently deplete OH levels to promote extended
SO2 lifetimes (e.g., Bekki, 1995).
4.1.6 Volcanic HCl emissionsACCEPTED into the UTLS MANUSCRIPT
A striking feature of Figure 4 that we highlight here is that MLS has detected injection of HCl into the UTLS by several explosive eruptions since 2004. We do not provide a detailed analysis of the MLS HCl data here, but rather point out the value of the MLS measurements (also including SO2 and potentially other halogen species such as CH3Cl; Table 1), which have not been widely used by the volcanological community (see also Pumphrey et al., 2015). HCl is an important halogen compound in the stratosphere. It is the dominant chlorine reservoir species
57 ACCEPTED MANUSCRIPT
and although it does not react directly with ozone, it is nevertheless a potential source of reactive halogen species (such as ClO) that destroy ozone, and hence any volcanic input of HCl into the stratosphere could be significant. Also notable in Figure 4 is an inverted ‗tape-recorder‘-like signal in the MLS HCl data (absent in the SO2 data), with higher HCl mixing ratios transported to lower altitudes in the northern hemisphere winter. Since the major source of stratospheric HCl is the UV photolysis of chlorine-containing compounds (e.g., CFCs) in the upper stratosphere
(note that Fig. 4 only extends to 47 hPa pressure), we assume that this pattern is due to tropical upwelling of HCl-poor air during the northern hemisphere summer, HCl production at higher altitudes, followed by HCl redistribution by the Brewer-Dobson circulation. The MLS HCl data at 147 hPa are noisy and although MLS HCl retrievals at pressures above 100 hPa are not recommended for scientific use (Livesey et al., 2011), the volcanic HCl signal in large eruption clouds is clearly strong enough to be detected at this level (Fig. 4).
Over the past few decades, opinions have varied on the potential for volcanic injection of
HCl (and other halogens) into the stratosphere. Symonds et al. (1988) estimated an annual global volcanic HCl flux of 0.4–11 Tg and suggested that <10% of these emissions originated from large explosive eruptions that would transmit them efficiently to the stratosphere. However, direct measurementsACCEPTED of stratospheric HCl after MANUSCRIPT major eruptions found relatively modest (1982 El
Chichón) or barely detectable (1991 Pinatubo) increases in HCl loadings above the stratospheric background (Mankin and Coffey, 1984; Mankin et al., 1992), although the observations were made weeks or months after the eruptions (and both eruptions produced measurable ozone depletion). Subsequent modeling work by Tabazadeh and Turco (1993) suggested that scavenging of volcanic HCl by condensed supercooled water in Plinian eruption columns could effectively remove more than 99% of emitted chlorine, precluding stratospheric HCl injection,
58 ACCEPTED MANUSCRIPT
whilst having minimal impact on less soluble SO2. However, Textor et al. (2003) pointed out flaws in the Tabazadeh and Turco (1993) study, namely that they used a relatively simple treatment of the dynamics, omitted some detailed cloud microphysical processes, and assumed that all hydrometeors were liquid (ice was later discovered to be dominant in some volcanic clouds; e.g., Rose et al., 1995). Accounting for the effects of ice particles, Textor et al. (2003) concluded that ice could scavenge 50-90% of HCl and 10-30% of SO2 in eruption columns, and predicted that >25% and 80% of emitted HCl and SO2, respectively, could potentially reach the stratosphere during explosive eruptions.
Recent satellite observations seem to confirm the predictions of Textor et al. (2003), and suggest that explosive eruptions frequently inject detectable HCl into the stratosphere (Fig. 4; e.g., Prata et al., 2007; Theys et al., 2014). Note that daily maximum SO2 and HCl mixing ratios retrieved by MLS are shown in Figure 4, and some weaker volcanic signals may be omitted, although the larger eruptions clearly stand out. Furthermore, due to the limited spatial sampling of MLS (section 3.5) some smaller eruptions may not be detected, or the regions of highest SO2 and HCl concentrations in volcanic clouds may not be sampled. Table 7 lists some HCl and SO2 mixing ratios measured by MLS in volcanic clouds since 2004, along with estimated HCl/SO2 mass ratios calculatedACCEPTED from collocated SO2 and MANUSCRIPT HCl vertical profiles. Actual volcanic SO2 and
HCl concentrations are likely to be underestimated by MLS since the vertical extent of volcanic plumes is typically much smaller than the MLS vertical resolution of ~3 km (Livesey et al., 2011;
Theys et al., 2014). Furthermore, since SO2 and HCl are retrieved using data from separate MLS radiometers with different fields of view, the averaging effects differ for the two gases: the vertical × horizontal (cross-track) × along-track resolution of MLS SO2 and HCl retrievals in the lower stratosphere is ~3×6×180 km and ~3×3×300-400 km, respectively (Livesey et al. 2011).
59 ACCEPTED MANUSCRIPT
Thus, since MLS averages volcanic SO2 and HCl concentrations over different atmospheric volumes, coupled with limited spatial sampling, we treat the HCl/SO2 ratios in Table 7 with caution. Nevertheless, some broad conclusions can be drawn. The HCl/SO2 ratios observed in volcanic eruption clouds (~0.01-0.03; Table 7) are around an order of magnitude lower than values (~0.1-0.3) estimated for other eruptions based on petrological data or in-situ measurements (e.g., Westrich and Gerlach, 1992; Gerlach et al., 1996; Rose et al., 2006). This may partly reflect uncertainties on the MLS-derived ratios, but would also be consistent with greater scavenging of HCl relative to SO2 in eruption plumes (Textor et al., 2003). Observed variations in HCl/SO2 ratios between eruptions may be a consequence of variable HCl/SO2 ratios in the source emissions (related to depth of degassing) or may reflect variable plume liquid water or total water content. Figure 4 shows a notable contrast in HCl signal between the 2008
Kasatochi and 2009 Sarychev Peak eruptions, which were similar in magnitude. Sampling issues notwithstanding, a possible explanation for the difference is that liquid water was more abundant in the Kasatochi eruption plume, since it erupted through a pre-existing crater lake (Waythomas et al., 2010), resulting in more effective scrubbing of HCl in the emissions. The 2011 Nabro eruption also produced a surprisingly strong UTLS HCl signal (Fig. 4, 10), given its non- subduction tectonic settingACCEPTED (northeast Afar, Eritrea) MANUSCRIPT where Cl enrichment from subducted seawater is absent. However, in this case the MLS-derived HCl/SO2 ratio of 0.03 is roughly commensurate with ratios measured in tropospheric plumes at other East African Rift volcanoes
(~0.04 at Erta ‗Ale and ~0.03 at Nyiragongo; Sawyer et al., 2008a, 2008b), perhaps suggesting minimal scavenging of HCl in the Nabro plume (or similar scavenging efficiencies of SO2 and
HCl).
60 ACCEPTED MANUSCRIPT
4.2 Passive volcanic SO2 degassing
We have shown above that the flux of SO2 from volcanic eruptions is well constrained, with the exception of lower magnitude events (VEI ≤2). However, there is more uncertainty on the global flux of SO2 emitted by passive (i.e., non-eruptive) degassing (e.g., Oppenheimer et al., 2011;
Shinohara, 2013), which also propagates into estimates of volcanic fluxes of other gases, such as
CO2 (e.g., Burton et al., 2014). The ground-based perspective on global volcanic degassing must be aggregated from diverse sources, including journal publications, volcano observatories, instrument networks (e.g., Galle et al., 2010), and Smithsonian GVP reports, and consequently takes considerable time to generate and update. Satellite measurements offer the significant advantage of providing a daily, global snapshot of volcanic SO2 emissions, collected by a single instrument using the same retrieval technique, and thus provide more timely and consistent assessments of the dominant volcanic SO2 sources at a given time.
Currently, OMI, GOME-2 and OMPS can provide daily observations of tropospheric volcanic SO2 plumes from space (e.g., Carn et al., 2007; Carn et al., 2008b; McCormick et al.,
2012, 2013; Carn et al., 2013; Yang et al., 2013), and the quality of measurements will improve with future sensors such as the Tropospheric Monitoring Instrument (TROPOMI; http://www.tropomi.euACCEPTED). Measurement noise andMANUSCRIPT variable spatial resolution limits the SO2 sources that can be detected in a single, daily satellite measurement to the strongest emitters
(OMI has greater sensitivity than GOME-2 or OMPS, due to higher spatial resolution), but spatial and temporal data averaging techniques can also be used to detect weaker SO2 sources
(e.g., Fioletov et al., 2011, 2013). The latter are less useful for volcano monitoring purposes, where timeliness is critical, but can be used to generate accurate, annual SO2 emissions inventories based on satellite observations. Here, we have used the technique described in
61 ACCEPTED MANUSCRIPT
section 3.4 to identify the most persistent volcanic SO2 sources detected by OMI in the 2004-
2014 period (Fig. 15).
Results of the daily OMI SO2 data analysis are shown in Figure 15. In these plots higher average values of the OSI generally indicate stronger SO2 emissions; hence the observations of
Anatahan and Nevado del Ruiz, for example, indicate shorter periods of strong SO2 degassing in
2004-2013. The SO2 emissions from four volcanoes (Ambrym, Kilauea, Popocatepetl and
Nyiragongo-Nyamulagira) have been notably more persistent than others in the ~10 years of satellite data, and SO2 emissions from Ambrym and Nyiragongo were also detected prior to 2004 in TOMS SO2 data (e.g., Carn, 2004). Three of these persistent SO2 sources host at least one active lava lake (Ambrym, Kilauea, and Nyiragongo-Nyamulagira), reflecting efficient and persistent SO2 outgassing from low-viscosity, basaltic magmas. Although many of the detected sources (e.g., Etna, Bagana, Nevado del Ruiz) figure prominently in previous volcanic SO2 emissions inventories (e.g., Andres and Kasgnoc, 1998), the global OMI measurements permit identification of sources in poorly monitored regions such as Papua New Guinea, Indonesia,
Vanuatu and Kamchatka, and also provide unique insight into the interannual variability of SO2 emissions (Fig. 15). For example, Sabancaya (Peru) and Sinabung (Indonesia) do not appear as strong SO2 sources basedACCEPTED on OMI measurements MANUSCRIPT in 2004-2013 (Fig. 15a), but are prominent in
2014 due to recently increased activity (Fig. 15b; Global Volcanism Program, 2014).
Although we do not convert the OMI SO2 measurements into SO2 emission rates here
(techniques for doing this are under development; e.g., Carn et al., 2013), recent compilations of ground-based SO2 measurements confirm high SO2 fluxes (exceeding ~10 kt/day at times) at
Ambrym, Kilauea, Popocatepetl, Nyiragongo, Etna and Bagana, amongst others (e.g.,
Oppenheimer et al., 2011; Elias and Sutton, 2012; Shinohara, 2013). We also note that several of
62 ACCEPTED MANUSCRIPT
these volcanoes (Nyiragongo, Popocatepetl, Ambrym, Etna) are regarded to be dominant sources of volcanic CO2 emissions (Burton et al., 2013). However, other sources with previously reported high SO2 emission rates, such as Láscar (N. Chile; 2.4 kt/day; Mather et al., 2004) are conspicuously absent from Figure 15, and this appears to be corroborated by more recent (post-
2004) ground-based SO2 measurements indicating substantially reduced SO2 emissions from
Láscar (~0.1-0.6 kt/day; Henney et al., 2012; Tamburello et al., 2014; Lopez et al., 2015). We therefore contend that UV satellite measurements of tropospheric SO2 plumes provide arguably the timeliest information on the status of global volcanic SO2 sources, especially considering that it is crucial to constrain the strongest sources in order to estimate global or regional SO2 fluxes
(e.g., Mori et al., 2013). We are not aware of any volcano with a reported SO2 emission rate of 1 kt/day or more since 2004 that has not been detected at some point in OMI SO2 measurements.
As SO2 fluxes from an increasing number of weaker volcanic SO2 sources are being measured using sensitive ground-based or airborne techniques (e.g., McGonigle et al., 2004; Mori et al.,
2006; Bani et al., 2013; Lopez et al., 2013b; Tamburello et al., 2014; Saing et al., 2014; Lopez et al., 2015; Smekens et al., 2015; Stebel et al., 2015), combining these new data with satellite observations will result in more accurate volcanic SO2 emissions inventories, encompassing a wider range of sourcesACCEPTED. MANUSCRIPT
Whilst the tropospheric SO2 plume detection method employed here (section 3.4) aims to omit major eruptive emissions (since eruptive SO2 clouds typically drift rapidly away from the source, beyond the 50 km search radius), SO2 plumes associated with continuous eruptive activity may still be captured. It is either known or likely that much of the SO2 detected at
Dukono, Anatahan, Kliuchevskoi, Zhupanovsky, Soufriere Hills, Sakura-jima and perhaps others is associated with continuous eruptions, i.e., the emissions are not passive in the strict sense
63 ACCEPTED MANUSCRIPT
(involving no coeval eruption of magma). Both Kilauea‘s and Bagana‘s SO2 emissions are also associated with ongoing, long-term effusive eruptions (Elias and Sutton, 2012; McCormick et al.,
2012) and hence could be regarded as eruptive discharges. Satellite observations of volcanic ash could potentially be used to distinguish plumes emitted by continuous eruptions, but ash measurements are typically less sensitive and noisier than SO2 measurements. Although of minor importance when attempting to estimate total volcanic SO2 fluxes to the atmosphere, the distinction between passive and eruptive degassing becomes important when assessing volcanic hazards or calculating volumes of degassed magma stored in the crust based on observed SO2 emissions.
To visualize the geographical distribution and temporal variation of the major tropospheric volcanic SO2 sources, the latitudes of the tropospheric SO2 plumes detected by OMI in 2004-2014 are shown in Figure 16 (note that some eruptive degassing is also apparent).
Persistent volcanic SO2 sources are manifested as continuous horizontal ‗bands‘ on this plot (Fig.
16): prominent near-continuous bands occur at ~20ºN (Kilauea and Popocatepetl; with increased
SO2 degassing apparent in March 2008 associated with the summit eruption at Kilauea; Elias and
Sutton, 2012); ~1-2ºS (Nyiragongo and Nyamuragira); ~4-9ºS (volcanoes of Indonesia and
Papua New Guinea);ACCEPTED and ~16ºS (Ambrym and MANUSCRIPT other Vanuatuan volcanoes such as Aoba and
Gaua). Other notable features of Figure 16 include elevated SO2 emissions from Anatahan
(~16ºN) in 2005-2008 with a subsequent decline in 2009-2014 (McCormick et al., 2015), increased SO2 emissions from Nevado del Ruiz (~5ºN) in early 2012 (Global Volcanism
Program, 2012), and at Turrialba (~10ºN) in 2009-2010 (Campion et al., 2012). There is also a striking contrast between the tropics and high latitudes (Fig. 16). This partly reflects the distribution of SO2 sources (most of the dominant volcanic SO2 sources are in the tropics) but is
64 ACCEPTED MANUSCRIPT
also a result of seasonal variations in the availability and sensitivity of UV measurements at high latitudes, and higher wind speeds at extratropical latitudes (which dilute SO2 column amounts in tropospheric plumes). Detection of degassing from Etna (~37ºN), Kamchatkan volcanoes (~50-
60ºN) and southern Chilean volcanoes (~35-40ºS) shows a clear seasonal dependence with wintertime observational gaps. However, the absence of a persistent SO2 degassing signal in
Iceland (~63-66ºN) is probably genuine, due to a lack of inter-eruptive SO2 emissions from
Icelandic volcanoes.
One conclusion that can be drawn from this analysis is that more ground-based measurements of SO2 emissions from high-latitude volcanoes (e.g., in Kamchatka, the Kurile
Islands and the Aleutian Islands), especially in winter, are required to improve volcanic gas emissions budgets. However, ground-based UV remote sensing techniques are also likely to be challenging in the relative darkness and adverse weather conditions of high-latitude winter, although IR spectroscopy and camera techniques have potential as they are not dependent on UV radiation (e.g., Lopez et al., 2013b; Prata and Bernardo, 2014; Lopez et al., 2015). Moderate- to high-resolution IR satellite retrievals of lower tropospheric SO2 (e.g., MODIS, ASTER) may also be possible for some plumes but routine detection of passive SO2 degassing is likely to be restricted by thermalACCEPTED contrast and sensitivity issues. MANUSCRIPT
5. Discussion
With at least 16 satellite instruments capable of measuring volcanic SO2 emissions currently in orbit (Table 1), the flow of SO2 observations from space has never been more profuse. This availability of remote sensing assets for detecting volcanic SO2 is set to continue for at least the next decade, as degrading instruments or those nearing the end of their operational lifetimes are
65 ACCEPTED MANUSCRIPT
supplanted by sensors such as TROPOMI, ensuring that the SO2 flux from volcanic eruptions will remain well constrained for the foreseeable future. Furthermore, should another Pinatubo- scale or larger eruption occur in the near future, current satellite measurements are capable of providing much higher quality data than they were in 1991. A Pinatubo-scale SO2 emission could potentially be tracked for weeks or months with the improved sensitivity of existing sensors relative to TOMS. Open questions to consider include whether an even larger SO2 release would saturate current SO2 retrieval algorithms during the initial emission, precluding accurate SO2 mass loading measurements until the cloud began to disperse, and how a persistent, widespread SO2 and sulfate aerosol veil from such a large event would impact remote sensing measurements during its atmospheric residence.
Although SO2 emissions are reasonably well constrained, a major unresolved question concerns the magnitude of global volcanic emissions of other sulfur species, most notably H2S, which have been previously estimated at 1.4-35 Tg/yr of S (e.g., Halmer et al., 2002;
Oppenheimer et al., 2011). Hydrogen sulfide can comprise up to ~40% of total sulfur gases (SO2
+ H2S) in high-temperature volcanic gas samples (Oppenheimer et al., 2011), and petrological evidence suggests that it may be dominant in some eruptions. Luhr et al. (1984) proposed that most of the sulfur gasesACCEPTED released by the 1982 ElMANUSCRIPT Chichón eruption were in the form of H2S, and
Hobbs et al. (1982) reported significant quantities of H2S in the May 1980 Mt St Helens eruption cloud. Although IASI is capable of detecting volcanic H2S from space (Table 1), the only reported observations to date are for the 2008 Kasatochi and 2011 Grimsvötn eruptions (Clarisse et al., 2011; Sigmarsson et al., 2013). These new observations of eruptive H2S emissions are valuable, though satellite measurements of H2S will never be as comprehensive as SO2
66 ACCEPTED MANUSCRIPT
measurements since the IR H2S absorption lines are weak and hence sensitivity is low (Clarisse et al., 2011).
There appears to be significant scope for more work on the rates of key reactions involved in H2S oxidation, but McKeen et al. (1984) suggest that H2S is oxidized to SO2 relatively quickly (~3 day chemical lifetime) provided OH is readily available. Hence the degree to which satellite measurements will capture H2S emissions converted to SO2 will depend on the relative rates of H2S oxidation to SO2 and SO2 conversion to sulfate, i.e., if H2S is oxidized relatively quickly, the SO2 produced will be detected and factored in to the total eruption sulfur budget. However, if OH concentrations were sufficient to promote efficient H2S oxidation, then it also likely that SO2 would be simultaneously oxidized. Oxidation of H2S has been invoked to explain increases in measured SO2 observed in the days following some explosive eruptions, such as the 1982 El Chichón and 1992 Mount Spurr eruptions (e.g., Bluth et al., 1995; Schneider et al., 1999; Rose et al., 2000, 2001), but such increases could also be reasonably attributed to other causes such as a reduction in volcanic cloud optical depth, reduced ash content, or sequestration of SO2 by ice in the young volcanic cloud (e.g., Rose et al., 2000, 2001), and are often subject to some ambiguity.
Although someACCEPTED H2S is likely to be present MANUSCRIPT in all volcanic eruption plumes, it may be safe to assume that SO2 dominates the sulfur inventory in emissions from volcanoes in subduction zone environments (where most of the major volcanic SO2 sources are located), where oxidized magmas with high SO2:H2S ratios are typical. Magmas in hot-spot and divergent plate contexts may also be relatively oxidized (Oppenheimer et al., 2011). Aiuppa et al. (2005) report a
SO2/H2S molar ratio of 20 in emissions from Etna in 2003, and H2S was not detected in emissions from Gorely volcano (Kamchatka) in 2011 (Aiuppa et al., 2012), although Edmonds et
67 ACCEPTED MANUSCRIPT
al. (2010) found a relatively significant quantities of H2S in gas emissions from SHV
(Montserrat) in July 2008 during an eruptive pause (SO2/H2S = 1.9-3.6). Emissions of H2S may be more significant during precursory activity at wet volcanoes, where phreatomagmatic eruptions may precede the main magmatic eruption phase, as was suggested for Kasatochi
(Clarisse et al., 2011). Clearly, however, further work is needed to better constrain volcanic H2S emissions on a global scale. Global S emissions from volcanic arcs have been estimated to be
~10-11 Tg/year (Hilton et al., 2002; Shinohara, 2013), equivalent to ~20-22 Tg/yr SO2.
Assuming ~16 Tg/yr SO2 emitted by passive degassing in volcanic arcs (Shinohara, 2013) and
~2 Tg/yr SO2 emitted by explosive eruptions (Table 3) yields a total of ~18 Tg/yr SO2 or 9 Tg/yr
S. Based on an estimated annual emission of 10-11 Tg S, this leaves a 1-2 Tg/yr S shortfall, which if emitted as H2S would be ~1.2-2.1 Tg/yr H2S.
One approach that could provide further insight into eruptive emissions of H2S is to identify significant discrepancies between petrological estimates of S in pre-eruptive magmatic gas phases (Scaillet and Pichavant, 2003; Scaillet et al., 2003) and observations of SO2 emissions. Such disagreement could indicate the discharge of substantial H2S emissions that were not detected in satellite data (or alternatively, rapid removal of SO2). One possible example is the April 1979 eruptionACCEPTED of Soufriere St Vincent, MANUSCRIPT for which Scaillet et al. (2003) predicted a total S release of ~0.5 Tg from the melt and gas phase (i.e., ~1 Tg SO2). However, only ~3 kt of
SO2 (0.3% of the predicted SO2 yield) was detected by TOMS following the eruption, despite adequate measurement conditions (a volcanic ash cloud was detected). Furthermore, the eruption has been linked to a large SAOD anomaly in 1979-80 (Fig. 13; Rosen and Hofmann, 1980), which would not be commensurate with such a low SO2 release. Taken together, this evidence
68 ACCEPTED MANUSCRIPT
may suggest that the bulk of the sulfur emitted by Soufriere St Vincent was in the form of H2S, which may have been oxidized sufficiently slowly to prevent any SO2 detection by TOMS.
Finally, although we do not discuss them in detail here, UV satellite measurements are also providing unique observations of reactive halogen species such as BrO and OClO in volcanic clouds (e.g., Afe et al., 2004; Theys et al., 2009; Rix et al., 2012; Hörmann et al., 2013;
Theys et al., 2014). Such measurements permit assessment of the potential impacts of volcanic eruptions on stratospheric ozone (e.g., Kutterolf et al., 2013). The HCl and SO2 profile measurements from MLS are also advantageous, since the impact of a volcanic eruption on the stratospheric ozone layer will depend primarily on the altitude of the volcanic emissions relative to the peak in the ozone concentration. Theys et al. (2014) found significant differences in the quality of volcanic BrO and OClO retrievals from available UV satellite instruments
(SCIAMACHY, OMI and GOME-2), with certain sensors optimized for particular volatile species, highlighting the fact that multiple instruments are required to ensure accurate measurements of all volcanic emissions. Another potentially interesting but as yet unexplored aspect of satellite measurements of reactive halogens in volcanic clouds is the different overpass times of the sensors. For example, we might expect systematic differences between BrO measurements by GOMEACCEPTED-2 (9:30 am overpass) MANUSCRIPT and OMI (1:45 pm overpass) due to photochemical effects on BrO production in volcanic plumes, although other factors such as spatial resolution and plume aerosol loading will also play a role.
6. Conclusions
In this contribution we have reviewed ~36 years of satellite measurements of volcanic SO2 emissions from 1978-2014. During this period, SO2 emissions from every significant volcanic
69 ACCEPTED MANUSCRIPT
eruption (VEI≥4) have been detected and quantified, providing accurate constraints on the mean volcanic SO2 flux to the UTLS (~1-2 Tg/yr) and allowing robust interpretation of trends in
SAOD. Measurements from multiple UV and IR satellite instruments are required to achieve optimum coverage, to ensure detection of eruptions at all latitudes and at night, and to reduce uncertainty on measured SO2 loadings for individual eruptions. Microwave sensors such as MLS can provide crucial complementary information on the altitude of volcanic SO2 and HCl injections. Although there is greater uncertainty on SO2 emissions from smaller eruptions (VEI
0-3), this SO2 is not expected to be significant for climate. New techniques being adopted for ground-based measurements of volcanic SO2 (e.g., UV and IR imaging cameras) will likely improve constraints on SO2 fluxes from small eruptions, which may be an important component of the tropospheric volcanic SO2 budget in years with low total SO2 emissions or no large eruptions. Furthermore, CO2 emissions from such activity may be more significant due to the longer atmospheric residence time of CO2 relative to SO2.
Reanalysis of the relationship between VEI and SO2 emissions with the inclusion of recent satellite measurements confirms a first-order relationship between eruption magnitude and
SO2 loading. However, due to significant variability in SO2 loading for individual VEI classes
(coupled with the inherentACCEPTED challenges in assigning MANUSCRIPT accurate VEIs), the use of VEI to assess the potential climate impact of eruptions in the remote sensing era is discouraged. Current satellite measurements can provide accurate and timely estimates of both SO2 loading and altitude, with this capability expected to continue in the next decade, obviating the need for any proxy for climate impact (but the VEI certainly remains a useful indicator of the overall significance of an eruption). Specification of eruption altitude is crucial for climate modeling, and one of the weaknesses of our current volcanic SO2 emissions database is that eruptions are characterized by
70 ACCEPTED MANUSCRIPT
a single altitude, whereas in reality the vertical profile of an explosive eruption column can be complex and variable. The assumption of a single SO2 injection altitude can result in significant errors in the UTLS SO2 budget, especially for complex, continuous eruptions (e.g., the 2010
Eyjafjallajökull, 2011 Nabro and 2014-15 Holuhraun eruptions), which will propagate into estimates of climate impacts. In this regard, the application of inverse modeling techniques to estimate the syn-eruptive vertical profile of SO2 emissions appears particularly powerful, and is facilitated by the availability of multiple UV and IR satellite measurements of volcanic SO2 emissions. Such techniques could be used to determine the peak injection altitude or multiple injection altitudes.
For the past decade, satellite instruments have also provided new constraints on passive volcanic degassing of SO2. A synthesis of ~10 years of OMI SO2 measurements reveals the most persistent volcanic SO2 sources detected from space, several of which have minimal or no ground-based monitoring currently in place. Our analysis demonstrates the key role of satellite observations in identifying the major volcanic SO2 sources at any given time. It has been postulated that a few dominant SO2 sources exert a major control over regional and global volcanic SO2 fluxes (e.g., Mori et al., 2013; Shinohara, 2013), and hence a combination of satellite SO2 measurementsACCEPTED of the major SO2 emittersMANUSCRIPT with the increasingly comprehensive ground-based SO2 data for volcanoes with lower SO2 emission rates will lead to vastly improved
SO2 emissions budgets. However, observations of SO2 emissions from high-latitude volcanoes
(e.g., in Kamchatka, the Kurile Islands and the Aleutian Islands) remain challenging, especially in winter.
Collation of a multi-decadal record of volcanic SO2 emissions allows us to draw some broad conclusions from the patterns that emerge. Volcanoes that exhibit minimal levels of
71 ACCEPTED MANUSCRIPT
2 passive or inter-eruptive SO2 degassing are responsible for sporadic (≥10 years), large (≥1 Tg)
SO2 emissions with potentially significant impacts on climate or SAOD (e.g., Pinatubo, El
Chichón, Kasatochi, Nabro); it is highly likely that the next, major SO2-producing eruption will occur at a volcano that has not erupted for centuries or more, where accumulation of a stored vapor phase may be occurring. We also note that such volcanoes may emit substantial amounts of CO2 in apparently quiescent periods, and more widespread ground-based or airborne CO2 flux measurements would be desirable; the challenge is to select appropriate targets from the many currently dormant volcanoes (active ground deformation may be one useful criterion). Perhaps more significant for climate on shorter timescales are the restless, volatile-rich systems exhibiting both passive and eruptive SO2 degassing (e.g., Rabaul, Manam, SHV, Ulawun) that produce smaller (VEI≤4) but frequent UTLS eruptions, which may have significant cumulative impacts on climate. For such systems, the magnitude of SO2 release is likely limited (≤1 Tg is typical) by near-continuous activity and bleeding of volatiles from resident magma at depth. This effect is even stronger for other systems where passive SO2 degassing dominates the total SO2 discharge and large eruptive SO2 emissions appear unlikely unless passive degassing is curtailed
(e.g., Popocatepetl, Tungurahua).
Although theACCEPTED SO2 data products from bothMANUSCRIPT UV and IR satellite instruments enjoy a long heritage and can be deemed relatively mature, some outstanding questions remain. Unraveling the causes of differences between UV and IR SO2 retrievals, or between IR retrievals in different spectral bands, found for some eruptions (e.g., Table 2) is needed to reduce the uncertainties on some eruption SO2 loadings. Such differences are most likely to originate from a complex combination of effects of SO2 altitude or vertical profile, overpass timing, plume aerosol loading
(e.g., ash, sulfate, hydrometeors), sensor spatial and spectral resolution, atmospheric water vapor,
72 ACCEPTED MANUSCRIPT
and total column ozone, and may be very challenging to untangle. Efforts should be made to determine the ‗optimum‘ satellite SO2 retrieval for each eruption. Validation of satellite SO2 measurements remains a challenge, particularly for the largest volcanic clouds, but may become more achievable using unmanned aerial vehicles (UAVs) or other rapidly-deployable in-situ measurement techniques. Finally, volcanic H2S emissions remain poorly constrained on a global scale and may be a significant fraction of the total S budget for some eruptions.
Although much can be achieved using satellite instruments deployed on multiple polar- orbiting platforms (and such observations are crucial for high latitude volcanoes), monitoring of volcanic SO2 emissions from space would greatly benefit from a global, geostationary capability, particularly in the UV. Currently, only the European SEVIRI and Japanese Advanced Himawari
Imager (AHI) sensors have IR channels capable of detecting SO2 emissions from geostationary orbit (Table 1). However, prospects for high temporal resolution UV monitoring of SO2 from geostationary or alternative orbits are improving. The recently deployed EPIC instrument on the
DSCOVR satellite at the L1 Earth-Sun Lagrange point will soon provide multiple daily UV SO2 retrievals for the sunlit Earth disk (including some coverage of high latitudes during the summer months). The SO2 sensitivity of EPIC will likely be intermediate between TOMS and OMI, permitting detection ACCEPTEDof large eruptions and strong MANUSCRIPT passive SO2 degassing. Later in the decade
(around 2017), the Tropospheric Emissions: Monitoring of Pollution (TEMPO) instrument will be launched on a geostationary satellite, providing UV observations of North America (including
Mexico). Since SO2 is a criteria pollutant affecting air quality, volcanic SO2 surveillance will indirectly benefit from efforts (such as the TEMPO mission) to measure anthropogenic SO2 emissions more frequently and track intercontinental transport of air pollution. TEMPO will form part of a global geostationary constellation for pollution monitoring, including UV
73 ACCEPTED MANUSCRIPT
instruments from Europe (Sentinel 4) and Asia (the Geostationary Environment Monitoring
Spectrometer [GEMS] being developed in Korea). We are therefore on the cusp of another major advance in our ability to measure volcanic SO2 emissions from space.
Acknowledgments
We acknowledge editors Tim Horscroft and Alessandro Aiuppa for inviting this review article.
NASA has provided generous support of satellite-based volcanic SO2 measurements, including recent grants NNX10AG60G (A Combined EOS Data and GEOS-Chem Modeling Study of the
Direct Radiative Forcing of Volcanic Sulfate Aerosois; PI: J. Wang), NNX11AK95G
(Continuation of Long-Term Sulfur Dioxide EDR with the SNPP/OMPS Nadir Mapper; PI: K.
Yang) and NNX13AF50G (Multi-Decadal Sulfur Dioxide Climatology from Satellite
Instruments; PI: N.A. Krotkov). We thank Jean-Paul Vernier (NASA Langley) for providing recent stratospheric aerosol data and Mauro Tomasi (ULB) for collating the data used to produce an earlier version of Figure 14. David Pyle and an anonymous reviewer provided thorough and constructive reviews of the paper. The article was written whilst the first author was on sabbatical at the Department of Mineral Sciences (DMS), National Museum of Natural History,
Smithsonian InstitutionACCEPTED, and SAC acknowledges MANUSCRIPT the Smithsonian for a visiting fellowship and the DMS staff for their hospitality. The volcanic SO2 emissions database described in this paper is available from the NASA Goddard Earth Sciences (GES) Data and Information Services
Center (DISC) as a level 4 MEaSUREs (Making Earth System Data Records for Use in Research
Environments) data product (Carn, 2015c).
74 ACCEPTED MANUSCRIPT
References
Afe, O.T., A. Richter, B. Sierk, F. Wittrock and J. P. Burrows (2004), BrO emission from
volcanoes - a survey using GOME and SCIAMACHY measurements, Geophys. Res.
Lett., 31, L24113, doi:10.1029/2004GL020994.
Aiuppa, A., Inguaggiato, S., McGonigle, A. J. S., O‘Dwyer, M., Op- penheimer, C., Padgett, M.
J., Rouwet, D., and Valenza, M. (2005), H2S fluxes from Mt. Etna, Stromboli, and
Vulcano (Italy) and implications for the sulfur budget at volcanoes, Geochim.
Cosmochim. Acta, 69, 1861–1871.
Aiuppa, A., G. Giudice, M. Liuzzo,G. Tamburello, P.Allard, S. Calabrese, I. Chaplygin,A. J. S.
McGonigle, and Y. Taran (2012), First volatile inventory for Gorely volcano,
Kamchatka, Geophys. Res. Lett., 39, L06307, doi:10.1029/2012GL051177.
Andres, R. J. and Kasgnoc, A. D. (1998), A time-averaged inventory of subaerial volcanic sulfur
emissions, J. Geophys. Res., 103, 25251– 25261.
Arfeuille, F., Weisenstein, D., Mack, H., Rozanov, E., Peter, T., and Brönnimann, S. (2014),
Volcanic forcing for climate modeling: a new microphysics-based data set covering years
1600–present, Clim. Past, 10, 359-375, doi:10.5194/cp-10-359-2014.
Bani, P., C. Oppenheimer,ACCEPTED V.I. Tsanev, S.A. Carn,MANUSCRIPT S.J. Cronin, R. Crimp, J.A. Calkins, D.
Charley, M. Lardy, and T.R. Roberts (2009), Surge in sulfur and halogen degassing from
Ambrym volcano, Vanuatu, Bull. Volcanol., 71(10), 1159-1168, doi:10.1007/s00445-
009-0293-7.
Bani, P., C. Oppenheimer, P. Allard, H. Shinohara, V. Tsanev, S. Carn, M. Lardy, and E.
Garaebeti (2012), First arc-scale volcanic SO2 budget for the Vanuatu archipelago, J.
Volcanol. Geotherm. Res., 211-212, 36-46, doi:10.1016/j.jvolgeores.2011.10.005.
75 ACCEPTED MANUSCRIPT
Bani, P., Surono, Hendrasto, M., Gunawan, H., and Primulyana, S. (2013), Sulfur dioxide
emissions from Papandayan and Bromo, two Indonesian volcanoes, Nat. Hazards Earth
Syst. Sci., 13, 2399-2407, doi:10.5194/nhess-13-2399-2013.
Barnie, T., D. Keir, I. Hamling, B. Hofmann, M. Belachew, S.A. Carn, D. Eastwell, J.O.S.
Hammond, A. Ayele, C. Oppenheimer, T. Wright (2015), Implications for patterns of
volcanism during rifting from the concluding dike in the Manda Hararo dike sequence,
In: Wright, T.J., A. Ayele, D.J. Ferguson, T. Kidane and C. Vye-Brown (eds) Magmatic
Rifting and Active Volcanism, Geol. Soc. Lon, Special Publications, 420 (in press).
Bauduin, S., L. Clarisse, C. Clerbaux, D. Hurtmans, and P.-F. Coheur (2014), IASI observations
of sulfur dioxide (SO2) in the boundary layer of Norilsk, J. Geophys. Res. Atmos., 119,
4253–4263, doi:10.1002/2013JD021405.
Bekki, S. (1995), Oxidation of volcanic SO2: a sink for stratospheric OH and H2O, Geophys. Res.
Lett., 22(8), 913-916.
Berresheim, H. and Jaeschke, W. (1983), The contribution of volcanoes to the global
atmospheric sulfur budget, J. Geophys. Res., 88, 3732–3740.
Bhartia, P.K., and C.W. Wellemeyer (2002), TOMS-V8 total O3 algorithm, version 2.0, ATBD-
OMI-02, NASAACCEPTED Goddard Space Flight MANUSCRIPT Cent., Greenbelt, Md. (Available at
http://eospso.gsfc.nasa.gov/eos_homepage/for_scientists/atbd/docs/OMI/ATBD-OMI-
02.pdf).
Blake, S. (2003), Correlations between eruption magnitude, SO2 yield, and surface cooling, In:
Volcanic Degassing (eds. C. Oppenheimer, D.M. Pyle and J. Barclay), Geological
Society, London, Special Publications, 213, pp.371-380.
76 ACCEPTED MANUSCRIPT
Bluth, G.J.S. and Carn, S.A. (2008), Exceptional sulfur degassing from Nyamuragira volcano,
1979- 2005, Int. J. Remote Sens., 29(22), 6667-6685.
Bluth, G. J. S., S. D. Doiron, S. C. Schnetzler, A. J. Krueger, and L. S. Walter (1992), Global
tracking of the SO2 clouds from the June, 1991 Mount Pinatubo eruptions, Geophys. Res.
Lett., 19, 151 – 154, doi:10.1029/91GL02792.
Bluth, G.J.S., C.C. Schnetzler, A.J. Krueger, and L.S. Walter (1993), The contribution of
explosive volcanism to global atmospheric sulfur dioxide concentrations, Nature, 366,
327-329.
Bluth, G.J.S., T.J. Casadevall, C.C. Schnetzler, S.D. Doiron, L.S. Walter, A. J. Krueger, and M.
Badruddin (1994), Evaluation of sulfur dioxide emissions from explosive volcanism: the
1982-83 eruptions of Galunggung, Java, Indonesia, J. Volcanol. Geotherm. Res., 63, 243-
256.
Bluth, G. J. S., C. J. Scott, I. E. Sprod, C. C. Schnetzler, A. J. Krueger, and L. S. Walter (1995),
Explosive emissions of sulfur dioxide from the 1992 Crater Peak eruptions, Mount Spurr
volcano, Alaska, in The 1992 eruptions of Crater Peak Vent, Mount Spurr volcano,
Alaska, edited by T. E. C. Keith, U.S. Geol. Surv. Bull., 2139, 37 – 45, US Government
Printing Office,ACCEPTED Washington, D.C. MANUSCRIPT
Bluth, G.J.S., W.I. Rose, I.E. Sprod, and A.J. Krueger (1997), Stratospheric loading of sulfur
from explosive volcanic eruptions, J. Geology, 105, 671-683.
Boichu, M., Menut, L., Khvorostyanov, D., Clarisse, L., Clerbaux, C., Turquety, S., and Coheur,
P.-F. (2013), Inverting for volcanic SO2 flux at high temporal resolution using
spaceborne plume imagery and chemistry-transport modelling: the 2010 Eyjafjallajökull
eruption case study, Atmos. Chem. Phys., 13, 8569-8584, doi:10.5194/acp-13-8569-2013.
77 ACCEPTED MANUSCRIPT
Bonny, E., S.A. Carn, F. Donnadieu, and M. Coltelli (2015), Paroxysmal degassing at Mt. Etna
in 2011-12, Earth Planet. Sci. Lett., in prep.
Bourassa, A. E., D. A. Degenstein, B. J. Elash, and E. J. Llewellyn (2010), Evolution of the
stratospheric aerosol enhancement following the eruptions of Okmok and Kasatochi:
Odin-OSIRIS measurements, J. Geophys. Res., 115, D00L03,
doi:10.1029/2009JD013274.
Bourassa, A. E., A. Robock, W. J. Randel, T. Deshler, L. A. Rieger, N. D. Lloyd, E. J.
Llewellyn, and D. A. Degenstein (2012), Large volcanic aerosol load in the stratosphere
linked to Asian monsoon transport, Science, 337(6090), 78–81,
doi:10.1126/science.1219371.
Bourassa, A. E., A. Robock, W. J. Randel, T. Deshler, L. A. Rieger, N. D. Lloyd, E. J.
Llewellyn, and D. A. Degenstein (2013), Response to comments on ―Large volcanic
aerosol load in the stratosphere linked to Asian monsoon transport‖, Science, 339(6120),
647, doi:10.1126/science.1227961.
Bovensmann, H., et al., (1999), SCIAMACHY - mission objectives and measurement modes, J.
Atmos. Sci., 56(2), 127-150.
Boynard, A., et al. (2014),ACCEPTED First simultaneous spaceMANUSCRIPT measurements of atmospheric pollutants in
the boundary layer from IASI: A case study in the North China Plain, Geophys. Res.
Lett., 41, 645–651, doi:10.1002/2013GL058333.
Brenot, H., Theys, N., Clarisse, L., van Geffen, J., van Gent, J., Van Roozendael, M., van der A,
R., Hurtmans, D., Coheur, P.-F., Clerbaux, C., Valks, P., Hedelt, P., Prata, F., Rasson, O.,
Sievers, K., and Zehner, C. (2014), Support to Aviation Control Service (SACS): an
78 ACCEPTED MANUSCRIPT
online service for near-real-time satellite monitoring of volcanic plumes, Nat. Hazards
Earth Syst. Sci., 14, 1099-1123, doi:10.5194/nhess-14-1099-2014.
Brühl, C., Lelieveld, J., Tost, H., Höpfner, M. and Glatthor, N. (2015), Stratospheric sulfur and
its implications for radiative forcing simulated by the chemistry climate model EMAC. J.
Geophys. Res. Atmos., 120: 2103–2118. doi: 10.1002/2014JD022430.
Burrows, J. P., et al., (1999), The Global Ozone Monitoring Experiment (GOME): mission
concept and first scientific results, J. Atmos. Sci., 56, 151-175.
Burton, M.R., G.M. Sawyer, and D. Granieri (2013), Deep carbon emissions from volcanoes,
Rev. Mineral. Geochem., 75, 323-354, doi:10.2138/rmg.2013.75.11.
Cadle, R. D. (1975), Volcanic emissions of halides and sulfur compounds to the troposphere and
stratosphere, J. Geophys. Res., 88, 1650–1652.
Campbell, J.R., Welton, E.J., Krotkov, N.A., Yang, K., Stewart, S.A., and Fromm, M.D. (2012),
Likely seeding of cirrus clouds by stratospheric Kasatochi volcanic aerosol particles near
a mid-latitude tropopause fold, Atmos. Environ., 46, 441-448.
Campion, R., G.G. Salerno, P.-F. Coheur, D. Hurtmans, L. Clarisse, K. Kazahaya, M. Burton, T.
Caltabiano, C. Clerbaux, and A. Bernard (2010), Measuring volcanic degassing of SO2 in
the lower troposphereACCEPTED with ASTER band MANUSCRIPT ratios, J. Volcanol. Geotherm. Res., 194, 42-54.
Campion, R., M. Martinez-Cruz, T. Lecoq, C. Caudron, J. Pacheco, G. Pinardi, C. Hermans, S.A.
Carn, and A. Bernard (2012), Space- and ground-based measurements of sulfur dioxide
emissions from Turrialba volcano (Costa Rica), Bull. Volcanol., 74(7), 1757-1770,
doi:10.1007/s00445-012-0631-z.
Carboni, E., Grainger, R., Walker, J., Dudhia, A., and Siddans, R. (2012), A new scheme for
sulphur dioxide retrieval from IASI measurements: application to the Eyjafjallajökull
79 ACCEPTED MANUSCRIPT
eruption of April and May 2010, Atmos. Chem. Phys., 12, 11417-11434, doi:10.5194/acp-
12-11417-2012.
Carn, S.A. (2004). Eruptive and passive degassing of sulfur dioxide at Nyiragongo volcano (D.R.
Congo): the 17 January 2002 eruption and its aftermath. Acta Vulcanologica, 14-15, 75-
86.
Carn, S.A. (2015a), Gas, Plume, and Thermal Monitoring. In: Sigurdsson, H., Houghton, B.,
Rymer, H., Stix, J., McNutt, S. (Eds.), The Encyclopedia of Volcanoes, 2nd Edition, pp.
1125–1149, Elsevier.
Carn, S.A. (2015b), On the detection and monitoring of effusive eruptions using satellite SO2
measurements, In: Harris, A.J.L., T. de Groeve, F. Garel and S.A. Carn (editors),
Detecting, Modeling and Responding to Effusive Eruptions, Geological Society of
London, Special Publications, 426, in press.
Carn, S.A. (2015c), Multi-Satellite Volcanic Sulfur Dioxide (SO2) Database Long-Term L4
Global, version 1, Goddard Earth Science Data and Information Services Center (GES
DISC), Greenbelt, MD, USA, Accessed Dec 22, 2015 at
doi:10.5067/MEASURES/SO2/DATA401 or
http://disc.sci.gsfc.nasa.gov/datacollection/MSVOLSO2L4_V1.htmlACCEPTED MANUSCRIPT.
Carn, S.A., and Bluth, G.J.S. (2003), Prodigious sulfur dioxide emissions from Nyamuragira
volcano, D.R. Congo. Geophys. Res. Lett., 30(23), 2211, doi:10.1029/2003GL018465.
Carn, S. A., and F. J. Prata (2010), Satellite-based constraints on explosive SO2 release from
Soufri re Hills Volcano, Montserrat, Geophys. Res. Lett., 37, L00E22, doi:10.1029/
2010GL044971.
80 ACCEPTED MANUSCRIPT
Carn, S.A. and T.M. Lopez (2011). Opportunistic validation of sulfur dioxide in the Sarychev
Peak volcanic eruption cloud, Atmos. Meas. Tech., 4, 1705-1712, doi:10.5194/amt-4-
1705-2011.
Carn, S.A., A.J. Krueger, G.J.S. Bluth, S.J. Schaefer, N.A. Krotkov, I.M. Watson, and S. Datta
(2003), Volcanic eruption detection by the Total Ozone Mapping Spectrometer (TOMS)
instruments: a 22-year record of sulfur dioxide and ash emissions. In: Volcanic
Degassing (eds. C. Oppenheimer, D.M. Pyle and J. Barclay), Geological Society, London,
Special Publications, 213, pp.177-202.
Carn, S.A., A.J. Krueger, N.A. Krotkov, and M.A. Gray (2004), Fire at Iraqi sulfur plant emits
SO2 clouds detected by Earth Probe TOMS, Geophys. Res. Lett., 31(19), L19105,
doi:10.1029/2004GL020719.
Carn, S. A., L. L. Strow, S. de Souza-Machado, Y. Edmonds, and S. Hannon (2005), Quantifying
tropospheric volcanic emissions with AIRS: The 2002 eruption of Mt. Etna (Italy),
Geophys. Res. Lett., 32, L02301, doi:10.1029/2004GL021034.
Carn, S.A., N.A. Krotkov, A.J. Krueger, K. Yang, and P.F. Levelt (2007), Sulfur dioxide
emissions from Peruvian copper smelters detected by the Ozone Monitoring Instrument,
Geophys. Res.ACCEPTED Lett., 34, L09801, doi:10.1029/2006GL029020. MANUSCRIPT
Carn, S.A., A. J. Prata, and S. Karlsdóttir (2008a), Circumpolar transport of a volcanic cloud
from Hekla (Iceland), J. Geophys. Res., 113, D14311, doi:10.1029/2008JD009878.
Carn, S.A., A.J. Krueger, N.A. Krotkov, S. Arellano, and K. Yang (2008b). Daily monitoring of
Ecuadorian volcanic degassing from space, J. Volcanol. Geotherm. Res., 176(1), 141-
150, doi:10.1016/j.jvolgeores.2008.01.029.
81 ACCEPTED MANUSCRIPT
Carn, S.A., J.S. Pallister, L. Lara, J.W. Ewert, S. Watt, A.J. Prata, R.J. Thomas, and G. Villarosa
(2009a), The unexpected awakening of Chaitén volcano, Chile, EOS Trans. AGU, 90(24),
205-206.
Carn, S.A., A.J. Krueger, N.A. Krotkov, K. Yang, and K. Evans (2009b), Tracking volcanic
sulfur dioxide clouds for aviation hazard mitigation, Nat. Hazards, 51(2), 325-343.
Carn, S.A., K. Froyd, B.E. Anderson, P. Wennberg, J. Crounse, K. Spencer, J. Dibb, N.A.
Krotkov, E.V. Browell, J.W. Hair, G. Diskin, G. Sachse, and S. Vay (2011). In-situ
measurements of tropospheric volcanic plumes in Ecuador and Colombia during TC4, J.
Geophys. Res., 116, D00J24, doi:10.1029/2010JD014718.
Carn, S.A., N.A. Krotkov, K. Yang, and A.J. Krueger (2013), Measuring global volcanic
degassing with the Ozone Monitoring Instrument (OMI), In: Pyle, D.M., Mather, T.A.
and Biggs, J. (eds) Remote Sensing of Volcanoes and Volcanic Processes: Integrating
Observation and Modeling, Geol. Soc. Lon, Special Publications, 380,
doi:10.1144/SP380.12.
Carn, S.A., K. Yang, A.J. Prata, and N.A. Krotkov (2015), Extending the long-term record of
volcanic SO2 emissions with the Ozone Mapping and Profiler Suite (OMPS) Nadir
Mapper, Geophys.ACCEPTED Res. Lett., 42, doi: 10.1002/2014GL062437MANUSCRIPT.
Chahine, M. T., et al. (2006), AIRS: Improving weather forecasting and providing new data on
greenhouse gases, Bull. Am. Meteorol. Soc., 87(7), 911-926, doi:10.1175/BAMS-87-7-
911.
Chin, M., and D.J. Jacob (1996), Anthropogenic and natural contributions to tropospheric
sulfate: A global model analysis, J. Geophys. Res., 101(D13), 18,691–18,699,
doi:10.1029/96JD01222.
82 ACCEPTED MANUSCRIPT
Christenson, B.W. (2000), Geochemistry of fluids associated with the 1995–1996 eruption of Mt.
Ruapehu, New Zealand: signatures and processes in the magmatic-hydrothermal system,
J. Volcanol. Geotherm. Res., 97, 1-30.
Christopher, T., M. Edmonds, M. C. S. Humphreys, and R. A. Herd (2010), Volcanic gas
emissions from Soufrière Hills Volcano, Montserrat 1995–2009, with implications for
mafic magma supply and degassing, Geophys. Res. Lett., 37, L00E04,
doi:10.1029/2009GL041325.
Clarisse, L., Coheur, P. F., Prata, A. J., Hurtmans, D., Razavi, A., Phulpin, T., Hadji-Lazaro, J.,
and Clerbaux, C. (2008), Tracking and quantifying volcanic SO2 with IASI, the
September 2007 eruption at Jebel at Tair, Atmos. Chem. Phys., 8, 7723-7734,
doi:10.5194/acp-8-7723-2008.
Clarisse, L., D. Hurtmans, A.J. Prata, F. Karagulian, and C. Clerbaux (2010), Retrieving radius,
concentration, optical depth, and mass of different types of aerosols from high-resolution
infrared nadir spectra, Appl. Optics, 49(19), 3713-3722.
Clarisse, L., Y. R'Honi, P.-F. Coheur, D. Hurtmans, and C. Clerbaux (2011a), Thermal infrared
nadir observations of 24 atmospheric gases, Geophys. Res. Lett., 38, L10802,
doi:10.1029/2011GL047271.ACCEPTED MANUSCRIPT
Clarisse, L., P.‐ F. Coheur, S. Chefdeville, J.‐ L. Lacour, D. Hurtmans, and C. Clerbaux
(2011b), Infrared satellite observations of hydrogen sulfide in the volcanic plume of the
August 2008 Kasatochi eruption, Geophys. Res. Lett., 38, L10804,
doi:10.1029/2011GL047402.
83 ACCEPTED MANUSCRIPT
Clarisse, L., D. Hurtmans, C. Clerbaux, J. Hadji-Lazaro, Y. Ngadi, and P.-F. Coheur (2012),
Retrieval of sulphur dioxide from the infrared atmospheric sounding interferometer
(IASI), Atmos. Meas. Tech., 5, 581–594, doi:10.5194/amt-5-581-2012.
Clarisse, L., Coheur, P.-F., Theys, N., Hurtmans, D., and Clerbaux, C. (2014), The 2011 Nabro
eruption, a SO2 plume height analysis using IASI measurements, Atmos. Chem. Phys., 14,
3095-3111, doi:10.5194/acp-14-3095-2014.
Clerbaux, C., P.-F. Coheur, L. Clarisse, J. Hadji-Lazaro, D. Hurtmans, S. Turquety, K. Bowman,
H. Worden, and S. A. Carn (2008), Measurements of SO2 profiles in volcanic plumes
from the NASA Tropospheric Emission Spectrometer (TES), Geophys. Res. Lett., 35,
L22807, doi:10.1029/2008GL035566.
Constantine, E. K., Bluth, G. J. S. and Rose, W. I. (2000), TOMS and AVHRR observations of
drifting volcanic clouds from the August 1991 eruptions of Cerro Hudson. In: Mouginis-
Mark, P. J., Crisp, J. A. & Fink, J. H. (eds) Remote Sensing of Active Volcanism,
Geophysical Monograph 116, AGU, Washington, DC, pp. 45-64.
Corradini, S., Merucci, L., and Prata, A. J. (2009), Retrieval of SO2 from thermal infrared
satellite measurements: correction procedures for the effects of volcanic ash, Atmos.
Meas. Tech., ACCEPTED2, 177-191, doi:10.5194/amt MANUSCRIPT-2-177-2009.
Corradini, S., L. Merucci, A. J. Prata, and A. Piscini (2010), Volcanic ash and SO2 in the 2008
Kasatochi eruption: Retrievals comparison from different IR satellite sensors, J. Geophys.
Res., 115, D00L21, doi:10.1029/2009JD013634.
Crutzen, P. (2006), Albedo enhancement by stratospheric sulfur injections: a contribution to
resolve a policy dilemma? Climatic Change, 77, 211-219.
84 ACCEPTED MANUSCRIPT
Deshler, T., R. Anderson-Sprecher, H. Jäger, J. Barnes, D.J. Hofmann, B. Clemesha, D.
Simonich, M. Osborn, R.G. Grainger, and S. Godin-Beekmann (2006), Trends in the
nonvolcanic component of stratospheric aerosol over the period 1971–2004, J. Geophys.
Res., 111, D01201, doi:10.1029/2005JD006089.
Diehl, T., A. Heil, M. Chin, X. Pan, D. Streets, M. Schultz, and S. Kinne (2012), Anthropogenic,
biomass burning, and volcanic emissions of black carbon, organic carbon, and SO2 from
1980 to 2010 for hindcast model experiments, Atmos. Chem. Phys. Discuss., 12, 24895-
24954, doi:10.5194/acpd-12-24895-2012.
Doeringer, D., A. Eldering, C. D. Boone, G. González Abad, and P. F. Bernath (2012),
Observation of sulfate aerosols and SO2 from the Sarychev volcanic eruption using data
from the Atmospheric Chemistry Experiment (ACE), J. Geophys. Res., 117, D03203,
doi:10.1029/2011JD016556.
Doiron, S.D., G.J.S. Bluth, C.C. Schnetzler, A.J. Krueger, and L.S. Walter (1991), Transport of
Cerro Hudson SO2 clouds, Eos Trans. AGU, 72(45), 489–498, doi:10.1029/90EO00354.
Draxler, R.R. and G.D. Rolph (2014), HYSPLIT (HYbrid Single-Particle Lagrangian Integrated
Trajectory) Model access via NOAA ARL READY Website
(http://ready.arl.noaa.gov/HYSPLIT.php).ACCEPTED MANUSCRIPT NOAA Air Resources Laboratory, Silver
Spring, MD.
Eckhardt, S., A. J. Prata, P. Seibert, K. Stebel, and A. Stohl (2008), Estimation of the vertical
profile of sulfur dioxide injection into the atmosphere by a volcanic eruption using
satellite column measurements and inverse transport modeling, Atmos. Chem. Phys., 8,
3881–3897, doi:10.5194/acpd-8-3881-2008.
85 ACCEPTED MANUSCRIPT
Edmonds, M., A. Aiuppa, M. Humphreys, R. Moretti, G. Giudice, R. S. Martin, R. A. Herd, and
T. Christopher (2010), Excess volatiles supplied by mingling of mafic magma at an
andesite arc volcano, Geochem. Geophys. Geosyst., 11, Q04005,
doi:10.1029/2009GC002781.
Eisinger, M., and J.P. Burrows (1998), Tropospheric sulfur dioxide observed by the ERS-2
GOME instrument, Geophys. Res. Lett., 25, 4177-4180.
Elias, T., and A.J. Sutton (2012), Sulfur dioxide emission rates from Kïlauea Volcano, Hawaii,
2007–2010, U.S. Geological Survey Open-File Report 2012–1107, 25 p, available at:
http://pubs.usgs.gov/of/2012/1107/.
Fairlie, T. D., Vernier, J.-P., Natarajan, M., and Bedka, K. M. (2014), Dispersion of the Nabro
volcanic plume and its relation to the Asian summer monsoon, Atmos. Chem. Phys., 14,
7045-7057, doi:10.5194/acp-14-7045-2014.
Fee, D., R.S. Matoza, K.L. Gee, T.B. Neilsen, and D.E. Ogden (2013), Infrasonic crackle and
supersonic jet noise from the eruption of Nabro Volcano, Eritrea, Geophys. Res. Lett., 40,
4199–4203, doi:10.1002/grl.50827.
Ferguson, D.J., T.D. Barnie, D.M. Pyle, C. Oppenheimer, G. Yirgu, E. Lewi, T. Kidane, S.A.
Carn, and I. HamlingACCEPTED (2010). Recent rift MANUSCRIPT-related volcanism in Afar, Ethiopia, Earth
Planet. Sci. Lett., 292(3-4), 409-418, doi:10.1016/j.epsl.2010.02.010.
Fioletov, V.E., Griffioen, E., Kerr, J., Wardle, D., and Uchino, O. (1998), Influence of volcanic
sulfur dioxide on spectral UV irradiance as measured by Brewer spectrophotometers,
Geophys. Res. Lett., 25, 1665–1668.
86 ACCEPTED MANUSCRIPT
Fioletov, V. E., C. A. McLinden, N. Krotkov, M. D. Moran, and K. Yang (2011), Estimation of
SO2 emissions using OMI retrievals, Geophys. Res. Lett., 38, L21811,
doi:10.1029/2011GL049402.
Fioletov, V. E., et al. (2013), Application of OMI, SCIAMACHY, and GOME-2 satellite SO2
retrievals for detection of large emission sources, J. Geophys. Res. Atmos., 118,
doi:10.1002/jgrd.50826.
Froidevaux, L., et al. (2008), Validation of Aura Microwave Limb Sounder HCl measurements,
J. Geophys. Res., 113, D15S25, doi:10.1029/2007JD009025.
Fromm, M., Nedoluha, G., and Charvát, Z. (2013), Comment on ―Large Volcanic Aerosol Load
in the Stratosphere Linked to Asian Monsoon Transport‖, Science, 339, 647,
doi:10.1126/science.1228605.
Fromm, M., G. Kablick III, G. Nedoluha, E. Carboni, R. Grainger, J. Campbell, and J. Lewis
(2014), Correcting the record of volcanic stratospheric aerosol impact: Nabro and
Sarychev Peak, J. Geophys. Res. Atmos., 119, 10,343–10,364,
doi:10.1002/2014JD021507.
Galle, B., M. Johansson, C. Rivera, Y. Zhang, M. Kihlman, C. Kern, T. Lehmann, U. Platt, S.
Arellano, andACCEPTED S. Hidalgo (2010), Network MANUSCRIPT for Observation of Volcanic and Atmospheric
Change (NOVAC)—A global network for volcanic gas monitoring: Network layout and
instrument description, J. Geophys. Res., 115, D05304, doi:10.1029/2009JD011823.
Gambacorta, A., and C.D. Barnet (2013), Methodology and information content of the NOAA
NESDIS operational channel selection for the Cross-track Infrared Sounder (CrIS), IEEE
Trans. Geosci. Remote Sens., 51(6), 3207-3216, doi: 10.1109/TGRS.2012.2220369.
87 ACCEPTED MANUSCRIPT
Gerlach TM (1991) Present-day CO2 emissions from volcanoes. EOS Trans, Am Geophys
Union 72 (23):249- 255.
Gerlach, T.M., H.R. Westrich, and R.B. Symonds (1996), Preeruption vapor in magma of the
climactic Mount Pinatubo eruption: source of the giant stratospheric sulfur dioxide cloud,
In: Newhall, C.G. and R.S Punongbayan (eds.), Fire and Mud: Eruptions and Lahars of
Mount Pinatubo, Philippines, University of Washington Press, Seattle, USA, pp. 415-
434.
Gerlach TM (2011) Volcanic versus anthropogenic carbon dioxide. EOS Trans, 92(24):201-208.
Global Volcanism Program (2008), Report on Kasatochi (United States). In: Wunderman, R
(ed.), Bulletin of the Global Volcanism Network, 33:7. Smithsonian Institution.
http://dx.doi.org/10.5479/si.GVP.BGVN200807-311130.
Global Volcanism Program (2012), Report on Nevado del Ruiz (Colombia). In: Wunderman, R
(ed.), Bulletin of the Global Volcanism Network, 37:8. Smithsonian Institution.
http://dx.doi.org/10.5479/si.GVP.BGVN201208-351020.
Global Volcanism Program (2013), Volcanoes of the World, v. 4.3.0. Venzke, E (ed.).
Smithsonian Institution, Washington, DC. Downloaded 19 Sept 2014.
http://dx.doi.org/10.5479/si.GVP.VOTW4ACCEPTED MANUSCRIPT-2013.
Global Volcanism Program (2014), Report on Sabancaya (Peru). In: Sennert, S K (ed.), Weekly
Volcanic Activity Report, 9 July-15 July 2014. Smithsonian Institution and US Geological
Survey.
Goitom, B., et al., (2015), First recorded eruption of Nabro volcano, Eritrea, 2011, Bull.
Volcanol. (in press).
88 ACCEPTED MANUSCRIPT
Graf, H.-F., Feichter, J., and Langmann, B. (1997), Volcanic sulfur emissions: estimates of
source strength and its contribution to the global sulfate distribution, J. Geophys. Res.,
102, 10727–10738.
Guo, S., G. J. S. Bluth, W. I. Rose, I. M. Watson, and A. J. Prata (2004), Re-evaluation of SO2
release of the 15 June 1991 Pinatubo eruption using ultraviolet and infrared satellite
sensors, Geochem. Geophys. Geosyst., 5, Q04001, doi:10.1029/2003GC000654.
Gurevich, G. S. and Krueger, A. J. (1997), Optimization of TOMS wavelength channels for
ozone and sulfur dioxide retrievals. Geophys. Res. Lett., 24(17), 2187-2190.
Halmer, M. M., Schmincke, H.-F., and Graf, H.-F. (2002), The annual volcanic gas input into the
atmosphere, in particular into the stratosphere: a global data set for the past 100 years, J.
Volcanol. Geotherm. Res., 115, 511–528.
Hammer, C.U. (1977), Past volcanism revealed by Greenland ice sheet impurities, Nature, 270,
482-486.
Hammer, C.U., Clausen, H.B., and Dansgaard, W. (1980), Greenland ice sheet evidence of
postglacial volcanism and its climatic impact, Nature, 288, 230-235.
Haywood J. M., Jones, A., Clarisse, L., Bourassa, A., Barnes, J., Telford, P., Bellouin, N.,
Boucher, O., ACCEPTEDAgnew, P., Clerbaux, C., MANUSCRIPTCoheur, P., Degenstein, D., and Braesicke, P.
(2010), Observations of the eruption of the Sarychev volcano and simulations using the
HadGEM2 climate model, J. Geophys. Res., 115, D21212, doi:10.1029/2010JD014447.
Head, E.M. (2005), A Remote Sensing Study of the Galapagos Islands Volcanoes Using TOMS
and HIRS/2 Satellite Sensors, MS Thesis, Michigan Technological University, Houghton,
MI, USA.
89 ACCEPTED MANUSCRIPT
Henney, L., Rodríguez, L., Watson, I. (2012), A comparison of SO2 retrieval techniques using
mini-UV spectrometers and ASTER imagery at Láscar volcano, Chile, Bull. Volcanol.
74, 589–594.
Hilton D. R., Fischer T. P. and Marty B. (2002) Noble gases and volatile recycling at subduction
zones. Rev. Mineral. 47, 319– 370.
Hobbs, P. V., J.P. Tuell, D. A. Hegg, L. F. Radke, and M. W. Eltgroth (1982), Particles and
gases in the emission from the 1980-1981 volcanic eruption of Mt. St. Helens, J.
Geophys. Res., 87, 11,062-11,086.
Hobbs, P., Radke, L., Lyons, J., Ferek, R. and Coffman, D. (1991), Airborne measurements of
particle and gas emissions from the 1990 volcanic eruptions of Mount Redoubt, J.
Geophys. Res., 96, 18735-18752.
Hofmann, D., J. Barnes, M. O‘Neill, M. Trudeau, and R. Neely (2009), Increase in background
stratospheric aerosol observed with lidar at Mauna Loa Observatory and Boulder,
Colorado, Geophys. Res. Lett., 36, L15808, doi:10.1029/2009GL039008.
Höpfner, M., Glatthor, N., Grabowski, U., Kellmann, S., Kiefer, M., Linden, A., Orphal, J.,
Stiller, G., von Clarmann, T., Funke, B., and Boone, C. D. (2013), Sulfur dioxide (SO2)
as observed byACCEPTED MIPAS/Envisat: temporal MANUSCRIPT development and spatial distribution at 15–45
km altitude, Atmos. Chem. Phys., 13, 10405-10423, doi:10.5194/acp-13-10405-2013.
Höpfner, M., Boone, C. D., Funke, B., Glatthor, N., Grabowski, U., Günther, A., Kellmann, S.,
Kiefer, M., Linden, A., Lossow, S., Pumphrey, H. C., Read, W. G., Roiger, A., Stiller, G.,
Schlager, H., von Clarmann, T., and Wissmüller, K. (2015), Sulfur dioxide (SO2) from
MIPAS in the upper troposphere and lower stratosphere 2002–2012, Atmos. Chem. Phys.
Discuss., 15, 5801-5847, doi:10.5194/acpd-15-5801-2015.
90 ACCEPTED MANUSCRIPT
Hörmann, C., Sihler, H., Bobrowski, N., Beirle, S., Penning de Vries, M., Platt, U., and Wagner,
T. (2013), Systematic investigation of bromine monoxide in volcanic plumes from space
by using the GOME-2 instrument, Atmos. Chem. Phys., 13, 4749-4781, doi:10.5194/acp-
13-4749-2013.
Hughes, E. J., L. C. Sparling, S. A. Carn, and A. J. Krueger (2012), Using horizontal transport
characteristics to infer an emission height time series of volcanic SO2, J. Geophys. Res.,
117, D18307, doi:10.1029/2012JD017957.
Ialongo, I., Hakkarainen, J., Kivi, R., Anttila, P., Krotkov, N.A., Yang, K., Li, C., Tukiainen, S.,
Hassinen, S., and Tamminen, J. (2015), Comparison of SO2 operational satellite products
with ground-based observations in northern Finland during the Icelandic Holuhraun
fissure eruption, Atmos. Meas. Tech., 8, 2279-2289, doi:10.5194/amt-8-2279-2015.
Kahn, R.A., and J. Limbacher (2012), Eyjafjallajökull volcano plume particle-type
characterization from space-based multi-angle imaging, Atmos. Chem. Phys., 12, 9459-
9477, doi:10.5194/acp-12-9459-2012.
Kahn, R. A., W.-H. Li, C. Moroney, D. J. Diner, J. V. Martonchik, and E. Fishbein (2007),
Aerosol source plume physical characteristics from space-based multiangle imaging, J.
Geophys. Res.ACCEPTED, 112, D11205, doi:10.1029/2006JD007647. MANUSCRIPT
Karagulian, F., L. Clarisse, C. Clerbaux, A. J. Prata, D. Hurtmans, and P. F. Coheur (2010),
Detection of volcanic SO2, ash, and H2SO4 using the Infrared Atmospheric Sounding
Interferometer (IASI), J. Geophys. Res., 115, D00L02, doi:10.1029/2009JD012786.
Karl, T.R., Arguez, A., Huang, B., Lawrimore, J.H., McMahon, J.R., Menne, M.J., Peterson,
T.C., Vose, R.S., and Zhang, H.M. (2015), Possible artifacts of data biases in the recent
global surface warming hiatus, Science, doi:10.1126/science.aaa5632.
91 ACCEPTED MANUSCRIPT
Kearney, C. S., Dean, K., Realmuto, V. J., Watson, I. M., Dehn, J. and Prata, F. (2008),
Observations of SO2 production and transport from Bezymianny volcano, Kamchatka
using the MODerate resolution Infrared Spectroradiometer (MODIS), Int. J. Remote
Sensing, 29(22), 6647-6665.
Kearney, C., I.M. Watson, G.J.S. Bluth, S.A. Carn, and V. Realmuto (2009), A Comparison of
Thermal Infrared and Ultraviolet Retrievals of SO2 in the Cloud Produced by the 2003
Al-Mishraq State Sulfur Plant Fire, Geophys. Res. Lett., 36, L10807,
doi:10.1029/2009GL038215.
Kellogg, W. W., Cadle, R. D., Allen, E. R., Lazrus, A. L., and Martell, E. A. (1972), The sulfur
cycle, Science, 175, 587–596.
Kerr, J.B., and W.J. Evans (1987), Comparison of ground-based and TOMS measurements of
SO2 from volcanic emissions, in: Scientific and Operational Requirements for TOMS
Data, edited by A.J. Krueger, pp. 60–69, NASA Conf. Publ. 2497, NASA, Washington,
D.C.
(http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19880004392_1988004392.pdf).
Khokhar, M. F., C. Frankenberg, M. Van Roozendael, S. Beirle, S. K hl, A. Richter, U. Platt,
and T. WagnerACCEPTED (2005), Satellite observations MANUSCRIPT of atmospheric SO2 from volcanic eruptions
during the time-period of 1996–2002, Adv. Space Res., 36, 879–887,
doi:10.1016/j.asr.2005.04.114.
Kratzmann, D.J., S. Carey, R. Scasso, and J.-A. Naranjo (2009), Compositional variations and
magma mixing in the 1991 eruptions of Hudson volcano, Chile, Bull. Volcanol., 71(4),
419-439.
92 ACCEPTED MANUSCRIPT
Kristiansen, N. I., et al. (2010), Remote sensing and inverse transport modeling of the Kasatochi
eruption sulfur dioxide cloud, J. Geophys. Res., 115, D00L16,
doi:10.1029/2009JD013286.
Kristiansen, N., A.J. Prata, A. Stohl, and S.A. Carn (2015), Stratospheric volcanic ash emissions
from the 13 February 2014 Kelut eruption, Geophys. Res. Lett., doi:
10.1002/2014GL062307 (in press).
Krotkov, N. A., S. A. Carn, A. J. Krueger, P. K. Bhartia, and K. Yang (2006), Band residual
difference algorithm for retrieval of SO2 from the Aura Ozone Monitoring Instrument
(OMI), IEEE Trans. Geosci. Remote Sens., 44(5), 1259– 1266,
doi:10.1109/TGRS.2005.861932.
Krotkov, N. A., et al. (2008), Validation of SO2 retrievals from the Ozone Monitoring Instrument
over NE China, J. Geophys. Res., 113, D16S40, doi:10.1029/2007JD008818.
Krotkov, N. A., M. R. Schoeberl, G. A. Morris, S. Carn, and K. Yang (2010), Dispersion and
lifetime of the SO2 cloud from the August 2008 Kasatochi eruption, J. Geophys. Res.,
115, D00L20, doi:10.1029/2010JD013984.
Krueger, A. J. (1983), Sighting of El Chichón sulfur dioxide clouds with the Nimbus 7 Total
Ozone MappingACCEPTED Spectrometer, Science MANUSCRIPT, 220, 1377–1379.
Krueger, A. J., Walter, L. S., Schnetzler, C. C. and Doiron, S. D. (1990), TOMS measurement of
the sulfur dioxide emitted during the 1985 Nevado del Ruiz eruptions. J. Volcanol.
Geotherm. Res., 41, 7-15.
Krueger, A. J., L. S. Walter, P. K. Bhartia, C. C. Schnetzler, N. A. Krotkov, I. Sprod, and G. J. S.
Bluth (1995), Volcanic sulfur dioxide measurements from the total ozone mapping
spectrometer instruments, J. Geophys. Res., D100, 14057–14076.
93 ACCEPTED MANUSCRIPT
Krueger, A. J., C. C. Schnetzler, and L. S. Walter (1996), The December 1981 eruption of
Nyamuragira volcano (Zaire), and the origin of the ―mystery cloud‖ of early 1982, J.
Geophys. Res., D101, 15191–15196.
Krueger, A. J., Schaefer, S. J., Krotkov, N., Bluth, G. and Barker, S. (2000), Ultraviolet remote
sensing of volcanic emissions. In: Mouginis-Mark, P. J., Crisp, J. A. & Fink, J. H. (eds)
Remote Sensing of Active Volcanism, Geophysical Monograph 116, AGU, Washington,
DC, pp. 25-43.
Krueger, A.J., N.A. Krotkov, and S.A. Carn (2008). El Chichon: the genesis of volcanic sulfur
dioxide monitoring from space, J. Volcanol. Geotherm. Res., 175, 408-414,
doi:10.1016/j.jvolgeores.2008.02.026.
Krueger, A. J., Krotkov, N., Yang, K., Carn, S., Vicente, G. and Schroeder, W. (2009),
Applications of satellite-based sulfur dioxide monitoring, IEEE JSTARS, 2, 293–298.
Kushendratno, J.S. Pallister, Kristianto, F.R. Bina, W. McCausland, S.A. Carn, N. Haerani, J.
Griswold, and R. Keeler (2012), Recent explosive eruptions and volcano hazards at
Soputan volcano – a basalt stratovolcano in north Sulawesi, Indonesia, Bull. Volcanol.,
74(7), 1581-1609.
Kutterolf, S., T. H. HansteenACCEPTED, K. Appel, A. Freundt MANUSCRIPT, K. Kr ger, W. Pérez, and H. Wehrmann
(2013), Combined bromine and chlorine release from large explosive volcanic eruptions:
A threat to stratospheric ozone?, Geology, 41(6), 707–710, doi:10.1130/G34044.1.
Lee, C., Richter, A., Weber, M., and Burrows, J. P. (2008), SO2 retrieval from SCIAMACHY
using the Weighting Function DOAS (WFDOAS) technique: comparison with Standard
DOAS retrieval, Atmos. Chem. Phys., 8, 6137-6145, doi:10.5194/acp-8-6137-2008.
94 ACCEPTED MANUSCRIPT
Lee, C., Martin, R. V. et al. (2009), Retrieval of vertical columns of sulfur dioxide from
SCIAMACHY and OMI: air mass factor algorithm development, validation, and error
analysis, J. Geophys. Res., 114, D22303, http://dx.doi.org/10.1029/2009JD012123.
Lee, C., Martin, R. V. et al. (2011), SO2 emissions and lifetimes: estimates from inverse
modeling using in situ and global, space based (SCIAMACHY and OMI) observations, J.
Geophys. Res, 116, D06304, http://dx.doi.org/10.1029/2010JD014758.
Le Guern, F. (1982), Les d bits de CO2 et de SO2 volcaniques dans l‘atmosph re, B. Volcanol.,
45, 197–202.
Levelt, P. F., G. H. J. van den Oord, M. R. Dobber, A. M lkki, H. Visser, J. de Vries, P.
Stammes, J. O. V. Lundell, and H. Saari (2006), The Ozone Monitoring Instrument, IEEE
Trans. Geosci. Remote Sens., 44(5), 1093–1101, doi:10.1109/TGRS.2006.872333.
Li, C., J. Joiner, N.A. Krotkov, and P.K. Bhartia (2013), A fast and sensitive new satellite SO2
retrieval algorithm based on principal component analysis: Application to the ozone
monitoring instrument, Geophys. Res. Lett., 40, doi:10.1002/2013GL058134.
Livesey, N.J., et al. (2011), Earth Observing System (EOS) Aura Microwave Limb Sounder
(MLS) Version 3.3 and 3.4 Level 2 data quality and description document., Tech. Rep.
JPL D-33509,ACCEPTED NASA Jet Propulsion Laboratory MANUSCRIPT, California Institute of Technology,
Pasadena, California, 91109-8099, available at: http://mls.jpl.nasa.gov/data/datadocs.php.
Lopez, T., S.A. Carn, C. Werner, D. Fee, P. Kelly, M. Doukas, M. Pfeffer, P. Webley, C. Cahill,
and D. Schneider (2013a), Evaluation of Redoubt Volcano's sulfur dioxide emissions by
the Ozone Monitoring Instrument, J. Volcanol. Geotherm. Res.,
doi:10.1016/j.jvolgeores.2012.03.002.
95 ACCEPTED MANUSCRIPT
Lopez, T., Fee, D., Prata, F., and Dehn, J. (2013b), Characterization and interpretation of
volcanic activity at Karymsky Volcano, Kamchatka, Russia, using observations of
infrasound, volcanic emissions, and thermal imagery. Geochem. Geophys. Geosyst., 14,
5106–5127.
Lopez, T., H.E. Thomas, A.J. Prata, A. Amigo, D. Fee, and D. Moriano (2015), Volcanic plume
characteristics determined using an infrared imaging camera, J. Volcanol. Geotherm.
Res., 300, 148-166, doi: 10.1016/j.jvolgeores.2014.12.009.
Luhr J.F., Carmichael I.S.E., and Varekamp J.C. (1984), The 1982 eruptions of El Chichòn
Volcano, Chiapas, Mexico: mineralogy and petrology of the anhydrite-bearing pumices,
J. Volcanol. Geotherm. Res., 23, 69-108.
Mankin, W.G., and M.T. Coffey (1984), Increased stratospheric hydrogen-chloride in the El
Chichon cloud, Science, 226, 170-172, doi:10.1126/science.226.4671.170.
Mankin, W.G., Coffey, M.T., and Goldman, A. (1992), Airborne observations of SO2, HCl, and
O3 in the stratospheric plume of the Pinatubo volcano in July 1991, Geophys. Res. Lett.,
19(2), 179–182, 1992.
Martin, D., Ardouin, B., Bergametti, G., Carbonnelle, J., Faivre-Pierret, R., Lambert, G., Le
Cloarec, M. F.ACCEPTED and Sennequier, G. (1986), MANUSCRIPT Geochemistry of sulfur in Mount Etna plume,
J. Geophys. Res., 91, 12249-12254.
Mart nez-Alonso, S., M. N. Deeter, H. M. Worden, C. Clerbaux, D. Mao, and J. C. Gille (2012),
First satellite identification of volcanic carbon monoxide, Geophys. Res. Lett., 39,
L21809, doi:10.1029/2012GL053275.
Mateshvili, N., Fussen, D., Mateshvili, G., Mateshvili, I., Vanhellemont, F., Kyrölä, E.,
Tukiainen, S., Kujanpää, J., Bingen, C., Robert, C., Tétard, C., and Dekemper, E. (2013),
96 ACCEPTED MANUSCRIPT
Nabro volcano aerosol in the stratosphere over Georgia, South Caucasus from ground-
based spectrometry of twilight sky brightness, Atmos. Meas. Tech., 6, 2563–2576,
doi:10.5194/amt-6-2563-2013.
Mather T.A., V.I. Tsanev, D.M. Pyle, A.J.S. McGonigle, C. Oppenheimer, and A.G. Allen
(2004), Characterization and evolution of tropospheric plumes from Lascar and Villarica
volcanoes, Chile, J. Geophys. Res., 109, D21303, doi:10.1029/2004JD004934.
Matiella Novak, A., I.M. Watson, H. Delgado-Granados, W.I. Rose, L. Cárdenas-González, and
V.J. Realmuto (2008), Volcanic emissions from Popocatépetl volcano, Mexico,
quantified using Moderate Resolution Imaging Spectroradiometer (MODIS) infrared
data: A case study of the December 2000–January 2001 emissions, J. Volcanol.
Geotherm. Res., 170(1-2), 76-85.
McCarthy, E.B., G.J.S. Bluth, I.M. Watson and A. Tupper (2008), Detection and analysis of the
volcanic clouds associated with the 18 and 28 August 2000 eruptions of Miyakejima
volcano, Japan, Int. J. Remote Sensing, 29(22), 6597-6620,
doi:10.1080/01431160802168400.
McGonigle, A. J. S., C. Oppenheimer, V. I. Tsanev, S. Saunders, K. Mulina, S. Tohui, J. Bosco,
J. Nahou, J. Kuduon,ACCEPTED and F. Taranu (2004), MANUSCRIPT Sulphur dioxide fluxes from Papua New
Guinea‘s volcanoes, Geophys. Res. Lett., 31, L08606, doi:10.1029/2004GL019568.
McGonigle, A. J. S., Delmelle, P., Oppenheimer, C., Tsanev, V. I., Delfosse, T., Williams-Jones,
G., Horton, K. and Mather, T. A. (2004), SO2 depletion in tropospheric volcanic plumes
Geophys. Res. Lett., 31, L13201.
97 ACCEPTED MANUSCRIPT
McCormick, B. T., Edmonds, M., Mather, T. A., and Carn, S. A. (2012), First synoptic analysis
of volcanic degassing in Papua New Guinea, Geochem. Geophys. Geosyst., 13, Q03008,
doi:10.1029/2011GC003945.
McCormick, B.T., M. Edmonds, T. Mather, R. Campion, C. Hayer, H.E. Thomas, and S.A. Carn
(2013), Volcano monitoring applications of the Ozone Monitoring Instrument (OMI), In:
Pyle, D.M., Mather, T.A. and Biggs, J. (eds) Remote Sensing of Volcanoes and Volcanic
Processes: Integrating Observation and Modeling, Geol. Soc. Lon, Special Publications,
380, doi:10.1144/SP380.11.
McCormick, B., C. Popp, B. Andrews, and E. Cottrell (2015), Ten years of satellite observations
reveal highly variable sulphur dioxide emissions at Anatahan volcano, Mariana Islands, J.
Geophys. Res. Atmos., 120, doi:10.1002/2014JD022856.
McKeen, S. A., S. C. Liu, and C. S. Kiang (1984), On the chemistry of stratospheric SO2 from
volcanic eruptions, J. Geophys. Res., 89(D3), 4873–4881,
doi:10.1029/JD089iD03p04873.
McPeters, R.D. (1993), The atmospheric SO2 budget for Pinatubo derived from NOAA-11
SBUV/2 spectral data, Geophys. Res. Lett., 20(18), 1971-1974.
McPeters, R D., D.F.ACCEPTED Heath, and B.M. Schlesinger MANUSCRIPT (1984), Satellite observation of SO2 from El
Chichon: Identification and Measurement, Geophys. Res. Lett., 11, 1203-1206.
Miles, G.M., R.G. Grainger, and E.J. Highwood (2004), The significance of volcanic eruption
strength and frequency for climate, Q. J. R. Meteorol. Soc., 130, 2361-2376.
Mori, T., K. Kazahaya, C. Oppenheimer, A.J.S. McGonigle, V. Tsanev, R. Olmos, M. Ohwada,
and T. Shuto (2006), Sulfur dioxide fluxes from the volcanoes of Hokkaido, Japan, J.
Volcanol. Geotherm. Res., 158, 235-243.
98 ACCEPTED MANUSCRIPT
Mori, T., H. Shinohara, K. Kazahaya, J.-I. Hirabayashi, T. Matsushima, T. Mori, M. Ohwada, M.
Odai, H. Iino, and M. Miyashita (2013), Time-averaged SO2 fluxes of subduction-zone
volcanoes: Example of a 32-year exhaustive survey for Japanese volcanoes, J. Geophys.
Res. Atmos., 118, 8662–8674, doi:10.1002/jgrd.50591.
Munro, R., M. Eisinger, C. Anderson, J. Callies, E. Corpaccioli, R. Lang, A. Lefebvre, Y.
Livschitz, and A. Perez Albinana (2006), GOME-2 on MetOp, paper presented at the
2006 EUMETSAT Meteorological Satellite Conference, Eur. Org. for the Exploit of
Meteorol. Satell., Helsinki.
Nagai, T., B. Liley, T. Sakai, T. Shibata, and O. Uchino (2010), Post-Pinatubo evolution and
subsequent trend of the stratospheric aerosol layer observed by mid-latitude lidars in both
hemispheres, SOLA, 6, 69–72, doi:10.2151/sola.2010-018.
Neely R. R. III., O. B. Toon, S. Solomon, J.-P. Vernier, C. Alvarez, J. M. English, K. H.
Rosenlof, M. J. Mills, C. G. Bardeen, J. S. Daniel, and J. P. Thayer (2013), Recent
anthropogenic increases in SO2 from Asia have minimal impact on stratospheric aerosol,
Geophys. Res. Lett., 40, doi:10.1002/grl.50263.
Newhall, C.G., and S. Self (1982), The volcanic explosivity index (VEI): An estimate of
explosivity mACCEPTEDagnitude for historic volcanism MANUSCRIPT, J. Geophys. Res., 87, 1231-1238.
Noel, S., Buchwitz, M., Bovensmann, H., Hoogen, R., and Burrows, J. P. (1999), Atmospheric
water vapor amounts retrieved from GOME satellite data, Geophys. Res. Lett., 26, 1841–
1844.
Noël, S., M. Buchwitz, H. Bovensmann, and J. P. Burrows (2002), Retrieval of total water
vapour column amounts from GOME/ERS-2 data, Adv. Space Res., 29(11), 1697-1702.
99 ACCEPTED MANUSCRIPT
Noël, S., M. Buchwitz, and J. P. Burrows (2004), First retrieval of global water vapour column
amounts from SCIAMACHY measurements, Atmos. Chem. Phys., 4, 111-125.
Noël, S., S. Mieruch, H. Bovensmann and J. P. Burrows (2008), Preliminary results of GOME-2
water vapour retrievals and first applications in polar regions, Atmos. Chem. Phys., 8,
1519-1529.
Nowlan, C.R., X. Liu, K. Chance, Z. Cai, T.P. Kurosu, C. Lee, and R.V. Martin (2011),
Retrievals of sulfur dioxide from the Global Ozone Monitoring Experiment 2
(GOME‐ 2) using an optimal estimation approach: Algorithm and initial validation, J.
Geophys. Res., 116, D18301, doi:10.1029/2011JD015808.
Oppenheimer, C., Francis, P. and Stix, J. (1998), Depletion rates of sulfur dioxide in tropospheric
volcanic plumes, Geophys. Res. Lett., 25, 2671-2674.
Oppenheimer, C., B. Scaillet, and R.S. Martin (2011), Sulfur degassing from volcanoes: source
conditions, surveillance, plume chemistry and Earth system impacts, Rev. Mineral.
Geochem., 73, 363-421, doi: 10.2138/rmg.2011.73.13.
Penning de Vries, M. J. M., Dörner, S., Puķīte, J., Hörmann, C., Fromm, M. D., and Wagner, T.
(2014), Characterisation of a stratospheric sulfate plume from the Nabro volcano using a
combination ACCEPTEDof passive satellite measurements MANUSCRIPT in nadir and limb geometry, Atmos. Chem.
Phys., 14, 8149-8163, doi:10.5194/acp-14-8149-2014.
Prata, A.J., and I.F. Grant (2001), Retrieval of microphysical and morphological properties of
volcanic ash plumes from satellite data: Application to Mt Ruapehu, New Zealand, Q. J.
R. Meteorol. Soc., 127, 2153-2179.
100 ACCEPTED MANUSCRIPT
Prata, A. J., and C. Bernardo (2007), Retrieval of volcanic SO2 column abundance from
Atmospheric Infrared Sounder data, J. Geophys. Res., 112, D20204,
doi:10.1029/2006JD007955.
Prata, A. J., and J. Kerkmann (2007), Simultaneous retrieval of volcanic ash and SO2 using
MSG- SEVIRI measurements, Geophys. Res. Lett., 34, L05813,
doi:10.1029/2006GL028691.
Prata, A. J. and C. Bernardo (2014), Retrieval of sulfur dioxide from a ground-based thermal
infrared imaging camera, Atmos. Meas. Tech., 7, 2807-2828, doi:10.5194/amt-7-2807-
2014.
Prata, A. J., S. Self, W. I. Rose, and D. M. O‘Brien (2003), Global, long-term sulphur dioxide
measurements from TOVS data: A new tool for studying explosive volcanism and
climate, in Volcanism and the Earth‘s Atmosphere, Geophys. Monogr. Ser., vol. 139,
edited by A. Robock and C. Oppenheimer, pp. 75–92, AGU, Washington, D. C.
Prata, A.J., S.A. Carn, A. Stohl and J. Kerkmann (2007), Long range transport and fate of a
stratospheric volcanic cloud from Soufrière Hills volcano, Montserrat, Atmos. Chem.
Phys., 7, 5093-5103.
Prata, A. J., G. Gangale,ACCEPTED L. Clarisse, and F. Karagulian MANUSCRIPT (2010), Ash and sulfur dioxide in the
2008 eruptions of Okmok and Kasatochi: Insights from high spectral resolution satellite
measurements, J. Geophys. Res., 115, D00L18, doi:10.1029/2009JD013556.
Prata, A. T., S. T. Siems, and M. J. Manton (2015), Quantification of volcanic cloud top heights
and thicknesses using A-train observations for the 2008 Chaitén eruption. J. Geophys.
Res. Atmos., 120, 2928–2950. doi: 10.1002/2014JD022399.
101 ACCEPTED MANUSCRIPT
Pumphrey, H. C., Read, W. G., Livesey, N. J., and Yang, K. (2015), Observations of volcanic
SO2 from MLS on Aura, Atmos. Meas. Tech., 8, 195-209, doi:10.5194/amt-8-195-2015.
Pugnaghi, S., G. Gangale, S. Corradini, and M. F. Buongiorno (2006), Mt. Etna sulfur dioxide
flux monitoring using ASTER‐ TIR data and atmospheric observations, J. Volcanol.
Geotherm. Res., 152, 74–90, doi:10.1016/j.jvolgeores.2005.10.004.
Pyle, D.M. (1995), Mass and energy budgets of explosive volcanic eruptions, Geophys. Res.
Lett., 22(5), 563-566.
Pyle, D.M. (2015), Sizes of Volcanic Eruptions, In: Sigurdsson, H., Houghton, B., Rymer, H.,
Stix, J., McNutt, S. (Eds.), The Encyclopedia of Volcanoes, 2nd Edition, pp. 257-264,
Elsevier.
Pyle, D.M., P.D. Beattie, and G.J.S. Bluth, (1996), Sulphur emissions to the stratosphere from
explosive volcanic eruptions, Bull. Volcanol., 57, 663-671.
Radke, L. F. (1982), Sulphur and sulphate from Mt Erebus, Nature, 299, 710-712.
Richter, A., F. Wittrock, M. Eisinger, and J. P. Burrows (1998), GOME observations of
tropospheric BrO in northern hemispheric spring and summer 1997, Geophys. Res. Lett.,
25, 2683-2686.
Rinsland, C.P., Gunson,ACCEPTED M.R., Ko, M.K.W., Weisenstein, MANUSCRIPT D.W., Zander, R., Abrams, M.C.,
Goldman, A., Sze, N.D., and Yue, G. K. (1995), H2SO4 photolysis: a source of sulfur
dioxide in the upper stratosphere, Geophys. Res. Lett., 22, 1109–1112,
doi:10.1029/95GL00917.
Read, W.G., L. Froidevaux, and J.W. Waters (1993), Microwave Limb Sounder (MLS)
measurements of SO2 from Mt. Pinatubo volcano, Geophys. Res. Lett., 20, 1299-1302,
doi:10.1029/93GL00831.
102 ACCEPTED MANUSCRIPT
Realmuto, V. J. (2000), The potential use of earth observing system data to monitor the passive
emission of sulfur dioxide from volcanoes. In: Mouginis-Mark, P.J., Crisp, J.A. & Fink,
J.H. (eds) Remote Sensing of Active Volcanism. Geophysical Monograph 116, AGU,
Washington, DC, pp. 101-115.
Realmuto, V.J. and Worden, H.M. (2000), The impact of atmospheric water vapor on the thermal
infrared remote sensing of volcanic sulfur dioxide emissions: A case study from the Pu'u
'O'o vent of Kilauea Volcano, Hawaii, J. Geophys. Res., 105, 21,497-21,508.
Realmuto, V.J., Abrams, M.J., Buongiornio, M.F., and Pieri, D.C. (1994), The use of
multispectral thermal infrared image data to estimate the sulfur dioxide flux from
volcanoes: a case study from Mount Etna, Sicily, 29 July 1986, J. Geophys. Res., 99,
481-488.
Realmuto, V.J., Sutton, A.J. and Elias, T. (1997), Multispectral thermal infrared imaging of
sulfur dioxide plumes: a case study from the East Rift Zone of Kilauea Volcano, Hawaii,
J. Geophys. Res., 102, 15057–15072.
Ridley, D. A., et al. (2014), Total volcanic stratospheric aerosol optical depths and implications
for global climate change, Geophys. Res. Lett., 41, 7763–7769,
doi:10.1002/2014GL061541.ACCEPTED MANUSCRIPT
Rix, M., Valks, P. et al. (2009), Satellite monitoring of volcanic sulfur dioxide emissions for
early warning of volcanic hazards, IEEE J-STARS, 2, 196–206.
Rix, M., P. Valks, N. Hao, D. Loyola, H. Schlager, H. Huntrieser, J. Flemming, U. Koehler, U.
Schumann, and A. Inness (2012), Volcanic SO2, BrO and plume height estimations using
GOME-2 satellite measurements during the eruption of Eyjafjallajökull in May 2010, J.
Geophys. Res., 117, D00U19, doi:10.1029/2011JD016718.
103 ACCEPTED MANUSCRIPT
Robock, A. (2000), Volcanic eruptions and climate, Rev. Geophys., 38, 191-219.
Robock, A., A. Marquardt, B. Kravitz, and G. Stenchikov (2009), Benefits, risks, and costs of
stratospheric geoengineering, Geophys. Res. Lett., 36, L19703,
doi:10.1029/2009GL039209.
Rodríguez, L. A., Watson, I. M., Edmonds, M., Ryan, G., Hards, V., Oppenheimer, C. M. and
Bluth, G. J. S. (2008), SO2 loss rates in the plume emitted by Soufrière Hills volcano,
Montserrat, J. Volcanol. Geotherm. Res., 173, 135 – 147.
Rolph, G.D. (2014), Real-time Environmental Applications and Display sYstem (READY)
Website (http://ready.arl.noaa.gov). NOAA Air Resources Laboratory, Silver Spring,
MD.
Rose, W. I., Delene, D. J., Schneider, D. J., Bluth, G. J. S., Krueger, A. J., Sprod, I., McKee, C.,
Davies, H. L., and Ernst, G. G. J. (1995), Ice in the 1994 Rabaul eruption cloud:
Implications for volcano hazard and atmospheric effects, Nature, 375, 477–479.
Rose, W.I., Bluth, G.J.S., and Ernst, G.G.J. (2000), Integrating retrievals of volcanic cloud
characteristics from satellite remote sensors—a summary, Phil. Trans. R. Soc. Lon. A
Math. Phys. Sci., 358, 1585–1606.
Rose, W.I., G.J.S. Bluth,ACCEPTED D.J. Schneider, G.G.J. MANUSCRIPT Ernst, C.M. Riley, L.J. Henderson, and R.G.
McGimsey (2001), Observations of volcanic clouds in their first few days of atmospheric
residence: the 1992 eruptions of Crater Peak, Mount Spurr Volcano, Alaska, J. Geology,
109, 677-694.
Rose, W.I., Gu, Y., Watson, I.M., Yu, T., Bluth, G.J.S., Prata, A.J., Krueger, A.J., Krotkov, N.,
Carn, S., Fromm, M.D., Hunton, D.E., Viggiano, A.A., Miller, T.M., Ballentin, J.O.,
Ernst, G.G.J., Reeves, J.M., Wilson, C. and Anderson, B.E. (2003), The February-March
104 ACCEPTED MANUSCRIPT
2000 eruption of Hekla, Iceland from a satellite perspective. In: Volcanism and the
Earth‘s Atmosphere (eds. A. Robock and C. Oppenheimer), AGU Geophysical
Monograph 139, pp. 107-132.
Rose, W. I., Millard, G. A., Mather, T. A., et al. (2006), The atmospheric chemistry of a 33–34
hour old volcanic cloud from Hekla Volcano (Iceland): Insights from direct sampling and
the application of chemical box modeling, J. Geophys. Res., 111, D20206,
doi:10.1029/2005JD006872.
Rosen, J.M., and D.J. Hofmann (1980), A stratospheric aerosol increase, Geophys. Res. Lett.,
7(9), 669-672.
Saing, U.B., P. Bani, and Kristianto (2014), Ibu volcano, a center of spectacular dacite dome
growth and long-term continuous eruptive discharges, J. Volcanol. Geotherm. Res., 282,
36-42.
Santer, B.D., et al. (2014), Volcanic contribution to decadal changes in tropospheric temperature,
Nature Geosci., 7, 185-189, doi:10.1038/ngeo2098.
Santer, B. D., et al. (2015), Observed multivariable signals of late 20th and early 21st century
volcanic activity, Geophys. Res. Lett., 42, 500–509, doi:10.1002/2014GL062366.
Sato, M., J. E. Hansen,ACCEPTED M. P. McCormick, and MANUSCRIPT J. B. Pollack (1993), Stratospheric aerosol optical
depth, 1850-1990, J. Geophys. Res., 98, 22987-22994.
Sawamura, P., Vernier, J., Barnes, J., Berkoff, T., Welton, E., Alados-Arboledas, L., Navas-
Guzmán, F., Pappalardo, G. A., Mona, L., Madonna, F., Lange, D., Sicard, M., Godin-
Beekmann, S., Payen, G., Wang, Z., Hu, S., Tripathi, S., Cordoba-Jabonero, C., and Hoff,
R. (2012), Stratospheric AOD after the 2011 eruption of Nabro volcano measured by
105 ACCEPTED MANUSCRIPT
lidars over the Northern Hemisphere, Environ. Res. Lett., 7, 034013, doi:10.1088/1748-
9326/7/3/034013.
Sawyer, G.M., S.A. Carn, C. Oppenheimer, V.I. Tsanez, and M. Burton (2008a), Investigation
into magma degassing at Nyiragongo volcano, Democratic Republic of Congo, Geochem.
Geophys. Geosyst., 9, Q02017, doi:10.1029/2007GC001829.
Sawyer, G.M., C. Oppenheimer, V.I. Tsanez, and G. Yirgu (2008b), Magmatic degassing at Erta
‗Ale volcano, Ethiopia, J. Volcanol. Geotherm. Res., 178, 837-846.
Scaillet B., Luhr, J.F., and Carroll, M.C. (2003), Petrological and volcanological constraints on
volcanic sulfur emissions to the atmosphere, In: Volcanism and the Earth‘s Atmosphere,
Robock A., and Oppenheimer C. (eds), Geophysical Monograph 139, 11-40, AGU,
Washington, DC.
Schmidt, G.A., D.T. Shindell, and K. Tsigaridis (2014), Reconciling warming trends, Nat.
Geosci., 7, 158–160.
Schneider, D.J., Rose, W.I., Coke, L.R., Bluth, G.J.S., Sprod, I.E., and Krueger, A.J. (1999),
Early evolution of a stratospheric volcanic eruption cloud as observed with TOMS and
AVHRR. J. Geophys. Res., 104, 4037–4050.
Schnetzler, C. C., Doiron,ACCEPTED S. D., Walter, L. S. MANUSCRIPTand Krueger, A. J. (1994), Satellite measurements
of sulfur dioxide from the Redoubt eruptions of 1989-1990. In: Miller, T. P. & Chouet, B.
A. (eds.) The 1989-1990 eruptions of Redoubt Volcano, Alaska. J. Volcanol. Geotherm.
Res., 62(1-4), 353-357.
Schnetzler, C.C., A.J. Krueger, G.S. Bluth, I.E. Sprod and L.S. Walter (1995), Comment on the
Paper ―The atmospheric SO2 budget for Pinatubo derived from NOAA-11 SBUV/2
spectral data‖ by R. D. McPeters, Geophys. Res. Lett., 22(3), 315-316.
106 ACCEPTED MANUSCRIPT
Schnetzler, C.C., G.J.S. Bluth, A.J. Krueger, and L.S. Walter (1997), A proposed volcanic sulfur
dioxide index (VSI), J. Geophys. Res., 102(B9), 20087-20091.
Schoeberl, M., S.D. Doiron, L.R. Lait, P.A. Newman, and A.J. Krueger (1993), A simulation of
the Cerro Hudson SO2 cloud, J. Geophys. Res., 98, 2949-2955.
Schwandner, F.M., and S.A. Carn (2011), Detecting volcanic CO2 from space: target mode
observations with GOSAT, Abstract A33C-0222, presented at 2011 Fall Meeting, AGU,
San Francisco, Calif., 5-9 Dec.
Schwandner, F.M., V.J. Realmuto, S.A. Carn, B. Kahn, T. Oda, A. Kuze, F. Kataoka, T. Krings,
P.J. Rayner, and K. Shiomi (2015), CO2 plume detection, verification, and flux
determination using OCO-2 data: volcanoes and power plants, Geophysical Research
Abstracts, 17, abstract EGU2015-7665, presented at EGU General Assembly 2015.
Sharma, K., S. Blake, S. Self, and A. J. Krueger (2004), SO2 emissions from basaltic eruptions,
and the excess sulfur issue, Geophys. Res. Lett., 31, L13612,
doi:10.1029/2004GL019688.
Sheng, J.-X., D. K. Weisenstein, B.-P. Luo, E. Rozanov, A. Stenke, J. Anet, H. Bingemer, and T.
Peter (2015), Global atmospheric sulfur budget under volcanically quiescent conditions:
Aerosol-chemistryACCEPTED-climate model predictions MANUSCRIPT and validation, J. Geophys. Res. Atmos.,
120, 256–276, doi:10.1002/2014JD021985.
Shibata, T. and Kouketsu, T. (2008), Volcanic Clouds from the 2007 Eruption of Jebel at Tair
(Yemen) Detected by Ground Based and Space Borne Lidar, Sci. Online Lett. Atmos., 4,
93-96, available at: https://www.jstage.jst.go.jp/article/sola/4/0/4_0_93/_pdf
107 ACCEPTED MANUSCRIPT
Shinohara, H. (2013), Volatile flux from subduction zone volcanoes: Insights from a detailed
evaluation of the fluxes from volcanoes in Japan, J. Volcanol. Geotherm. Res., 268, 46-
63.
Siebert, L., T. Simkin, and P. Kimberly (2010), Volcanoes of the World, 3rd Ed., University of
California Press, Berkeley and Los Angeles, CA.
Sigmarsson, O., B. Haddadi, S.A. Carn, S. Moune, J. Gudnason, K. Yang, and L. Clarisse
(2013), The sulfur budget of the 2011 Grímsvötn eruption, Iceland, Geophys. Res. Lett.,
40(23), 6095-6100, doi:10.1002/2013GL057760.
Smekens, J.-F., A.B. Clarke, M.R. Burton, A. Harijoko, and H.E. Wibowo (2015), SO2 emissions
at Semeru volcano, Indonesia: Characterization and quantification of persistent and
periodic explosive activity, J. Volcanol. Geotherm. Res., doi:
10.1016/j.jvolgeores.2015.01.006 (in press).
Smets, B., et al., (2014), Detailed multidisciplinary monitoring reveals pre- and co-eruptive
signals at Nyamulagira volcano (North Kivu, D.R.C.), Bull. Volcanol., 76(1), doi:
10.1007/s00445-013-0787-1.
Solomon, S., J.S. Daniel, R.R. Neely III, J.-P. Vernier, E.G. Dutton, and L.W. Thomason (2011),
The persistentlyACCEPTED variable ―background‖ MANUSCRIPT stratospheric aerosol layer and global climate
change, Science, 333, 866-870.
Spinei, E., S. A. Carn, N. A. Krotkov, G. H. Mount, K. Yang, and A. Krueger (2010), Validation
of ozone monitoring instrument SO2 measurements in the Okmok volcanic cloud over
Pullman, WA, July 2008, J. Geophys. Res., 115, D00L08, doi:10.1029/2009JD013492.
Spiro, P. A., Jacob, D. J., and Logan, J. A. (1992), Global Inventory of Sulfur Emissions with 1º
× 1º Resolution, J. Geophys. Res., 97, 6023–6036.
108 ACCEPTED MANUSCRIPT
Stebel, K., Amigo, A., Thomas, H., Prata, A.J. (2015), First estimates of fumarolic SO2 fluxes
from Putana volcano, Chile, using an ultraviolet imaging camera, Journal of Volcanology
and Geothermal Research, doi:10.1016/j.jvolgeores.2014.12.021 (in press).
Steinman, B.A., M.E. Mann, and S.K. Miller (2015), Atlantic and Pacific multidecadal
oscillations and Northern Hemisphere temperatures, Science, 347, 988-991, doi:
10.1126/science.1257856.
Stoiber, R. E., Williams, S. N., and Huebert, B. (1987), Annual contribution of sulfur dioxide to
the atmosphere by volcanoes, J. Volcanol. Geoth. Res., 33, 1–8.
Stothers, R.B., J.A. Wolff, S. Self, and M.R. Rampino (1986), Basaltic fissure eruptions, plume
heights, and atmospheric aerosols, Geophys. Res. Lett., 13(8), 725–728.
Surono, P. Jousset, J. Pallister, M. Boichu, M.F. Buongiorno, A. Budisantoso, F. Costa, S.
Andreastuti, F. Prata, D. Schneider, L. Clarisse, H. Humaida, S. Sumarti, C. Bignami, J.
Griswold, S.A. Carn, C. Oppenheimer and F. Lavigne (2012). The 2010 explosive
eruption of Java‘s Merapi volcano – a ‗100-year‘ event, J. Volcanol. Geotherm. Res.,
241-242, 121-135, doi:10.1016/j.jvolgeores.2012.06.018.
Symonds, R.B., W.I. Rose and M.H. Reed (1988), Contribution of Cl- and F-bearing gases to the
atmosphere byACCEPTED volcanoes, Nature, 334, MANUSCRIPT 415-418, doi:10.1038/334415a0.
Tabazedeh, A. and Turco, R. P. (1993), Stratospheric chlorine injection by volcanic eruptions:
HCl scavenging and implications for ozone, Science, 260, 1082–1086.
Tamburello, G., T. H. Hansteen, S. Bredemeyer, A. Aiuppa, and F. Tassi (2014), Gas emissions
from five volcanoes in northern Chile and implications for the volatiles budget of the
Central Volcanic Zone, Geophys. Res. Lett., 41, 4961–4969,
doi:10.1002/2014GL060653.
109 ACCEPTED MANUSCRIPT
Teggi, S., M.P. Bogliolo, M.F. Buongiorno, S. Pugnaghi, and A. Sterni (1999), Evaluation of
SO2 emissions from Mt. Etna using diurnal and nocturnal multispectral IR and visible
imaging spectrometer thermal infrared remote sensing images and radiative transfer
models, J. Geophys. Res., 104, 20,069-20,079.
Telling, J.W., V.J.B. Flower, and S.A. Carn (2015), A multi-sensor satellite assessment of SO2
emissions from the 2012-13 eruption of Tolbachik volcano, Kamchatka, J. Volcanol.
Geotherm. Res., in press.
Theys, N., M. Van Roozendael, B. Dils, F. Hendrick, N. Hao, and M. De Mazière (2009), First
satellite detection of volcanic bromine monoxide emission after the Kasatochi eruption,
Geophys. Res. Lett., 36, L03809, doi:10.1029/2008GL036552.
Theys, N., Campion, R., Clarisse, L., Brenot, H., van Gent, J., Dils, B., Corradini, S., Merucci,
L., Coheur, P.-F., Van Roozendael, M., Hurtmans, D., Clerbaux, C., Tait, S., and
Ferrucci, F. (2013), Volcanic SO2 fluxes derived from satellite data: a survey using OMI,
GOME-2, IASI and MODIS, Atmos. Chem. Phys., 13, 5945-5968, doi:10.5194/acp-13-
5945-2013.
Theys, N., I. De Smedt, M. Van Roozendael, L. Froidevaux, L. Clarisse, and F. Hendrick (2014),
First satellite ACCEPTEDdetection of volcanic OClO MANUSCRIPT after the eruption of Puyehue-Cord n Caulle,
Geophys. Res. Lett., 41, 667–672, doi:10.1002/2013GL058416.
Thomas, W., T. Erbertseder, T. Ruppert, M. van Roozendael, J. Verdebout, D. Balis, C. Meleti,
and C. Zerefos (2005), On the retrieval of volcanic sulfur dioxide emissions from GOME
backscatter measurements, J. Atmos. Chem., 50, 295-320.
Thomas, H.E., and I.M. Watson (2009), Observations of volcanic emissions from space: current
and future perspectives, Nat. Hazards, 54(2), 323-354, doi:10.1007/s11069-009-9471-3
110 ACCEPTED MANUSCRIPT
Thomas, H.E., I.M Watson, C. Kearney, S.A. Carn, and S.J. Murray (2009). A multi-sensor
comparison of sulphur dioxide emissions from the 2005 eruption of Sierra Negra
volcano, Galápagos Islands. Remote Sens. Environ., 113(6), 1331-1342,
doi:10.1016/j.rse.2009.02.019.
Thomas, H.E., I.M. Watson, S.A. Carn, A.J. Prata, and V.J. Realmuto (2011), A comparison of
AIRS, MODIS and OMI sulphur dioxide retrievals in volcanic clouds, Geomatics Nat.
Haz. Risk, 2(3), 217-232, doi:10.1080/19475705.2011.564212.
Tulet, P. and Villeneuve, N. (2011), Large scale modeling of the transport, chemical
transformation and mass budget of the sulfur emitted during the April 2007 eruption of
Piton de la Fournaise, Atmos. Chem. Phys., 11, 4533-4546, doi:10.5194/acp-11-4533-
2011.
Tupper, A., and K. Kinoshita (2003), Satellite, Air and Ground Observations of Volcanic Clouds
over Islands of the Southwest Pacific, South Pacific Study, 23(2), 21-46.
Tupper, A., and R. Wunderman (2009), Reducing discrepancies in ground and satellite-observed
eruption heights, J. Volcanol. Geotherm. Res., 186, 22-31.
Tupper, A., et al., (2004). An evaluation of volcanic cloud detection techniques during recent
significant eruptionsACCEPTED in the western ‗Ring MANUSCRIPT of Fire‘. Remote Sensing of Environment 91,
27–46. doi:10.1016/j.rse.2004.02.004.
Tupper, A., I. Itikarai, M. Richards, A.J. Prata, S.A. Carn, and D. Rosenfeld (2007). Facing the
challenges of the International Airways Volcano Watch: the 2004/05 eruptions of
Manam, Papua New Guinea, Weather and Forecasting 22(1), 175-191.
111 ACCEPTED MANUSCRIPT
Tupper, A., C. Textor, M. Herzog, H.-F. Graf and M.S. Richards (2009), Tall clouds from small
eruptions: the sensitivity of eruption height and fine ash content to tropospheric
instability, Nat. Hazards, 51, 375-401.
Uchino, O., Sakai, T., Nagai, T., Nakamae, K., Morino, I., Arai, K., Okumura, H., Takubo, S.,
Kawasaki, T., Mano, Y., Matsunaga, T., and Yokota, T. (2012), On recent (2008–2012)
stratospheric aerosols observed by lidar over Japan, Atmos. Chem. Phys., 12, 11975-
11984, doi:10.5194/acp-12-11975-2012.
Urai, M., (2004), Sulfur dioxide flux estimation from volcanoes using Advanced Spaceborne
Thermal Emission and Reflection Radiometer — a case study of the Miyakejima
volcano, Japan, J. Volcanol. Geotherm. Res., 134, 1–13.
Vaughan, M.A., K.A. Powell, D.M. Winker, C.A. Hostetler, R.E. Kuehn, W.H. Hunt, B.J.
Getzewich, S.A. Young, Z. Liu, and M.J. McGill (2009), Fully Automated Detection of
Cloud and Aerosol Layers in the CALIPSO Lidar Measurements, J. Atmos. Oceanic
Technol., 26, 2034–2050.
Vernier, J. P. et al. (2011), Major influence of tropical volcanic eruptions on the stratospheric
aerosol layer during the last decade, Geophys. Res. Lett., 38(12), L12,807.
Vernier, J.P., L.W. Thomason,ACCEPTED T.D. Fairlie, P. MANUSCRIPT Minnis, R. Palikonda, and K. M. Bedka (2013),
Comment on ―Large volcanic aerosol load in the stratosphere linked to Asian monsoon
transport‖, Science, 339(6120), 647, doi:10.1126/science.1227817. von Glasow, R., Bobrowski, N., and Kern, C. (2009), The effects of volcanic eruptions on
atmospheric chemistry, Chem. Geol., 263, 131-142.
Wadge, G., and L. Burt (2011), Stress field control of eruption dynamics at a rift volcano:
Nyamuragira, D.R. Congo, J. Volcanol. Geotherm. Res., 207, 1-15.
112 ACCEPTED MANUSCRIPT
Wallace P. J. (2005) Volatiles in subduction zone magmas: concentrations and fluxes based on
melt inclusion and volcanic gas data. J. Volcanol. Geotherm. Res. 140, 217–240.
Wallace, L., and W. Livingston (1992), The effect of the Pinatubo cloud on hydrogen chloride
and hydrogen fluoride, Geophys. Res. Lett., 19, 1209-1211.
Wallace P. J. and Carmichael I. S. E. (1992) Sulfur in basaltic magmas. Geochim. Cosmochim.
Acta 56, 1863–1874.
Wallace P.J., and Gerlach T.M. (1994), Magmatic vapor source for sulfur dioxide released
during volcanic eruptions: evidence from Mount Pinatubo, Science, 265, 497-499.
Wallace P. J. and Edmonds M. (2011) The sulfur budget in magmas: evidence from melt
inclusions, submarine glasses, and volcanic gas emissions. Rev. Mineral. Geochem., 73,
215–246.
Walter, D., K.-P. Heue, A. Rauthe-Schöch, C.A.M. Brenninkmeijer, L.N. Lamsal, N.A. Krotkov,
and U. Platt (2012), Flux calculation using CARIBIC DOAS aircraft measurements: SO2
emission of Norilsk, J. Geophys. Res., 117, D11305, doi:10.1029/2011JD017335.
Wang, J., Park, S., Zeng, J., Ge, C., Yang, K., Carn, S., Krotkov, N., and Omar, A. H. (2013),
Modeling of 2008 Kasatochi volcanic sulfate direct radiative forcing: assimilation of
OMI SO2 plumeACCEPTED height data and comparison MANUSCRIPT with MODIS and CALIOP observations,
Atmos. Chem. Phys., 13, 1895-1912, doi:10.5194/acp-13-1895-2013.
Waters, J.W., et al. (2006), The Earth Observing System Microwave Limb Sounder (EOS MLS)
on the Aura satellite, IEEE T. Geosci. Remote Sens., 44(5), 1075–1092.
Watson, I.M., Realmuto, V.J., Rose, W.I., Prata, A.J., Bluth, G.J.S., Gu, Y., Bader, C.E. and Yu,
T. (2004), Thermal infrared remote sensing of volcanic emissions using the moderate
resolution imaging spectroradiometer, J. Volcanol. Geotherm. Res., 135, 75−89.
113 ACCEPTED MANUSCRIPT
Waythomas, C. F., W. E. Scott, S. G. Prejean, D. J. Schneider, P. Izbekov, and C. J. Nye (2010),
The 7–8 August 2008 eruption of Kasatochi Volcano, central Aleutian Islands, Alaska, J.
Geophys. Res., 115, B00B06, doi:10.1029/2010JB007437.
Westrich H.R., and Gerlach T.M. (1992), Magmatic gas source for the stratospheric SO2 cloud
from the June 15, 1991 eruption of Mount Pinatubo, Geology, 20, 867-70.
Wigley, T. M. L. (2006), A combined mitigation/geoengineering approach to climate
stabilization, Science, 314, 452 – 454, doi:10.1126/science.1131728.
Winker, D.M., M.A. Vaughan, A. Omar, Y. Hu, K.A. Powell, Z. Liu, W.H. Hunt, and S.A.
Young (2009), Overview of the CALIPSO Mission and CALIOP Data Processing
Algorithms, J. Atmos. Oceanic Technol., 26, 2310–2323.
Winker, D. M., Z. Liu, A. Omar, J. Tackett, and D. Fairlie (2012), CALIOP observations of the
transport of ash from the Eyjafjallajökull volcano in April 2010, J. Geophys. Res., 117,
D00U15, doi:10.1029/2011JD016499.
Wright, R., Carn, S.A., and Flynn, L. (2005). A satellite chronology of the May 2003 eruption of
Anatahan volcano. J. Volcanol. Geotherm. Res., 146, 102-116.
Yang, K., N.A. Krotkov, A.J. Krueger, S.A. Carn, P.K. Bhartia, and P.F. Levelt (2007), Retrieval
of large volcanicACCEPTED SO2 columns from the MANUSCRIPT Aura Ozone Monitoring Instrument: Comparison
and limitations, J. Geophys. Res., 112, D24S43, doi:10.1029/ 2007JD008825.
Yang, K., N. A. Krotkov, A. J. Krueger, S. A. Carn, P. K. Bhartia, and P. F. Levelt (2009a),
Improving retrieval of volcanic sulfur dioxide from backscattered UV satellite
observations, Geophys. Res. Lett., 36, L03102, doi:10.1029/2008GL036036.
114 ACCEPTED MANUSCRIPT
Yang, K., X. Liu, N.A. Krotkov, A.J. Krueger and S.A. Carn (2009b). Estimating the altitude of
volcanic sulfur dioxide plumes from space-borne hyper-spectral UV measurements,
Geophys. Res. Lett., 36, L10803, doi:10.1029/2009GL038025.
Yang, K., X. Liu, P.K. Bhartia, N.A. Krotkov, S.A. Carn, E. Hughes, A.J. Krueger, R. Spurr and
S. Trahan (2010), Direct retrieval of sulfur dioxide amount and altitude from spaceborne
hyper-spectral UV measurements: theory and application, J. Geophys. Res., 115,
D00L09, doi:10.1029/2010JD013982.
Yang, K., R.R. Dickerson, S.A. Carn, C. Ge, and J. Wang (2013), First observations of SO2 from
the satellite Suomi NPP OMPS: Widespread air pollution events over China, Geophys.
Res. Lett., 40, doi:10.1002/grl.50952.
ACCEPTED MANUSCRIPT
115 ACCEPTED MANUSCRIPT
Tables
Table 1. Volcanic gases measured or potentially detectable from space
Volatile species
Cl
2 3
2
O S 2
Sensora 2 Timespan Reference(s)b
H CO CO SO H HCl BrO OClO CH TOMS* 1978-2005 1, 2 SBUV* (P) 1978-present 3, 4 HIRS* 1978-present 5 GOME 1995-2003 6, 7, 8 MODIS* 1999-present 9, 10 ASTER 1999-present 11, 12, 13 MOPITT 1999-present 14 SCIAMACHY (L) 2002-2012 8, 15, 16 MIPAS (L) 2002-2012 17 AIRS 2002-present 18, 19 ACE (L) 2003-present 20 SEVIRI 2004-present 21 OMI 2004-present 16, 22, 23, 24 MLS* (L) 1991-2001; 2004-present 25, 26, 27, 28 TES (P) 2004-present 29 GOME-2* 2006-present 16, 30, 31 IASI* 2006-present 14, 32, 33 OMPS* 2011-present 34 VIIRS 2011-present 35 CrIS 2011-present 36 AHI 2015-present 37 GOSAT (P) 2009-present 38 OCO-2 2014-present 39 Red = detected in a volcanic cloud; Light grey = potentially detectable but not yet proven in a volcanic context and/or not viable for routine volcanic measurements (e.g., due to background interference). a. Sensor acronyms: TOMS:ACCEPTED Total Ozone Mapping MANUSCRIPT Spectrometer; SBUV: Solar Backscatter Ultraviolet Instrument; HIRS: High-resolution Infrared Radiation Sounder; GOME: Global Ozone Monitoring Experiment; MODIS: Moderate resolution Imaging Spectroradiometer; ASTER: Advanced Spaceborne Thermal Emission and Reflection Radiometer; MOPITT: Measurements of Pollution in the Troposphere; SCIAMACHY: Scanning Imaging Absorption Spectrometer for Atmospheric Chartography; MIPAS: Michelson Interferometer for Passive Atmospheric Sounding; AIRS: Atmospheric Infrared Sounder; ACE: Atmospheric Chemistry Experiment; SEVIRI: Spinning Enhanced Visible and Infrared Imager; OMI: Ozone Monitoring Instrument; MLS: Microwave Limb Sounder; TES: Tropospheric Emission Spectrometer; IASI: Infrared Atmospheric Sounding Interferometer; OMPS: Ozone Mapping and Profiler Suite; VIIRS: Visible Infrared Imaging Radiometer Suite; CrIS: Cross-track Infrared Sounder; AHI: Advanced Himawari Imager; GOSAT: Greenhouse Gases Observing Satellite; OCO-2: Orbiting Carbon Observatory 2. * = Sensor flown (or to be flown) on multiple satellites; P: nadir profiling and/or pointable instrument (limited mapping capability); L: limb instrument (vertical profiling); all other instruments provide some spatial mapping capability (SCIAMACHY provided nadir and limb observations).
116 ACCEPTED MANUSCRIPT
b. 1. Krueger et al. (1995); 2. Krueger et al. (2000); 3. McPeters et al. (1984); 4. McPeters (1993); 5. Prata et al. (2003); 6. Eisinger and Burrows (1998); 7. Thomas et al. (2005); 8. Afe et al. (2004); 9. Watson et al. (2004); 10. Corradini et al. (2009); 11. Urai (2004); 12. Pugnaghi et al. (2006); 13. Campion et al. (2010); 14. Martínez-Alonso et al. (2012); 15. Lee et al. (2008); 16. Theys et al. (2014); 17. Höpfner et al. (2013); 18. Carn et al. (2005); 19. Prata and Bernardo (2007); 20. Doeringer et al. (2012); 21. Prata and Kerkmann (2007); 22. Krotkov et al. (2006); 23. Yang et al. (2007); 24. Carn et al. (2013); 25. Read et al. (1993); 26. Prata et al. (2007); 27. Pumphrey et al. (2015); 28. W.G. Read, pers. comm.; 29. Clerbaux et al. (2008); 30. Rix et al. (2009); 31. Theys et al. (2009); 32. Clarisse et al. (2008); 33. Clarisse et al. (2011b); 34. Carn et al. (2015); 35. V.J. Realmuto, pers. comm.; 36: Gambacorta and Barnet (2013); 37. Carn (2015a); 38. Schwandner and Carn (2011); 39. Schwandner et al. (2015).
ACCEPTED MANUSCRIPT
117 ACCEPTED MANUSCRIPT
Table 2. SO2 mass loadings for selected volcanic eruptions (1978-2014)
Total SO2 Volcano Eruption date(s) loading Sensor(s) Reference (Tg) Nyamuragira 1979-2004 ~25a TOMS Bluth and Carn (2008) Hekla 17 Aug 1980 0.6±0.1 TOMS Carn et al. (2008a) El Chichón 28 Mar 1982 1.6 TOMS Krueger et al. (2008) 3-4 Apr 1982 5.9 TOMS Krueger et al. (2008) Galunggung Apr 1982 – Jan 1983 2.03 TOMS Bluth et al. (1994) Nevado del Ruiz 13 Nov 1985 0.66 TOMS Krueger et al. (1990) Redoubt 15 Dec 1989 0.18 TOMS Schnetzler et al. (1994) Hekla 17 Jan 1991 0.4±0.1 TOMS Carn et al. (2008a) Pinatubo 15 Jun 1991 18±4 TOMS Guo et al. (2004) 19±4 HIRS/2 Guo et al. (2004) 17 MLS Read et al. (1993) 12-15 SBUV/2 McPeters (1993) Hudson 8-12 Aug 1991 4 TOMS Constantine et al. (2000) Mt. Spurr 27 Jun 1992 0.25 TOMS Bluth et al. (1995) 18 Aug 1992 0.3 TOMS Bluth et al. (1995) 17 Sep 1992 0.25 TOMS Bluth et al. (1995) Ruapehu 17 Jun 1996 0.08 HIRS/2 Prata et al. (2003) Hekla 26 Feb 2000 0.2 TOMS/HIRS Rose et al. (2003) Miyakejima 18 Aug 2000 0.043 Multiple McCarthy et al. (2008) Popocatepetl 23 Jan 2001 0.017 MODIS 7.3 µm Matiella Novak et al. (2008) Nyiragongo 17 Jan 2002 0.03 TOMS Carn (2004) Etna 30 Oct 2002 0.025 AIRS Carn et al. (2005) 0.022 AIRS Prata et al. (2007) Anatahan 10-12 May 2003 0.11 TOMS Wright et al. (2005) Soufriere Hills 13 Jul 2003 0.10 TOMS Carn and Prata (2010) 0.13 MODIS 7.3 µm Carn and Prata (2010) 0.12 AIRS Carn and Prata (2010) Soputan Oct 2004 – Oct 2008 0.11b OMI Kushendratno et al. (2012) Manam 28 Jan 2005 0.14 TOMS This study ACCEPTED0.14 MANUSCRIPTOMI This study 0.13 AIRS Prata and Bernardo (2007) Anatahan 6 Apr 2005 0.070 TOMS This study 0.075 OMI This study 0.077 AIRS Prata and Bernardo (2007) Sierra Negra 22 Oct 2005 1.74 OMI Yang et al. (2009a) 0.33 TOMS Thomas et al. (2009) 1.27 MODIS 8.6 µm Thomas et al. (2009) Karthala 24 Nov 2005 0.08 SEVIRI Prata and Kerkmann (2007) Soufriere Hills 20 May 2006 0.22 OMI Carn and Prata (2010) 0.23 AIRS Carn and Prata (2010) 0.18 SEVIRI Prata et al. (2007) P. Fournaise 4-9 Apr 2007 0.23 OMI Tulet and Villeneuve (2011) Manda Hararo 12 Aug 2007 0.03 OMI Ferguson et al. (2010) Jebel al-Tair 30 Sep 2007 0.08±0.02 Multiple Eckhardt et al. (2008)
118 ACCEPTED MANUSCRIPT
0.05 IASI Clarisse et al. (2008) Chaitén 2-8 May 2008 0.01 OMI Carn et al. (2009a) Okmok 12 Jul 2008 0.12 OMI Spinei et al. (2010) 0.3 AIRS Prata et al. (2010) 0.09 IASI This study Kasatochi 7 Aug 2008 1.7 IASI Karagulian et al. (2010) 1.2 AIRS Prata et al. (2010) 1.7 Multiple Kristiansen et al. (2010) 2.2 OMI Krotkov et al. (2010) 0.9±0.2 MODIS 7.3 µm Corradini et al. (2010) 2.7±0.8 MODIS 8.6 µm Corradini et al. (2010) 1.6 IASI Clarisse et al. (2012) 1.6 GOME-2 Nowlan et al. (2011) Dalafilla 3 Nov 2008 0.23 IASI This study Redoubt 23 Mar – 12 Jun 2009 0.54-0.61c OMI Lopez et al. (2013a) Sarychev Peak 15-16 Jun 2009 1.2 IASI Haywood et al. (2010) Manda Hararo 28 Jun 2009 0.03 OMI Ferguson et al. (2010) Eyjafjallajökull 1-12 May 2010 0.17 IASI+Model Boichu et al. (2013) 5-18 May 2010 1.2±0.5 GOME-2 Rix et al. (2012) 7 May 2010 0.18 IASI Carboni et al. (2012) Merapi 26 Oct – 4 Nov 2010 0.44 Multiple Surono et al. (2012) Kizimen 13-31 Dec 2010 0.06 IASI This study Kirishima 26 Jan 2011 0.03 IASI This study Grimsvötn 21 May 2011 0.38 OMI+IASI Sigmarsson et al. (2013) Cordon Caulle 4-6 Jun 2011 0.2 IASI Theys et al. (2013) Nabro 12-28 Jun 2011 4.5 Multiple Theys et al. (2013) Nyamuragira 6 Nov – 6 Dec 2011 ~1 OMI Theys et al. (2013) Tolbachik 27 Nov 2012 0.09 IASI This study Nov 2012 – Aug 2013 0.2 Multiple Telling et al. (2015) Copahue 22 Dec 2012 0.5 IASI This study Paluweh 2 Feb 2013 0.03 Multiple Carn et al. (2015) Kelut 14 Feb 2014 0.2 OMI This study 0.19 IASI This study
a. Total SO2 releaseACCEPTED for 14 effusive eruptions MANUSCRIPT b. Total SO2 release for 9 explosive eruptions c. Total SO2 release for several UTLS injections
119 ACCEPTED MANUSCRIPT
Table 3. Total annual eruptive volcanic SO2 emissions, 1978-2014
Year Total SO2 Explosive SO2 Explosive/Total emissions (kt) emissions (kt) SO2 (%) 1978a 87 87 100 1979 2945 63 2.1 1980 2882 1228 43 1981 5562 1430 26 1982 11033 9950 90 1983 348 260 75 1984 3957 252 6.4 1985 1241 852 69 1986 2086 872 42 1987 38 10 26 1988 580 300 52 1989 2329 177 7.6 1990 186 186 100 1991 26082 24214 93 1992 810 810 100 1993 450 450 100 1994 1874 360 19 1995a TOMS data gap 1996a 987 100 10 1997 41 41 100 1998 3265 38 1.2 1999 130 85 65 2000 653 336 51 2001 1783 122 6.8 2002 2626 271 10 2003 679 679 100 2004 2997 410 14 2005 4634 2501 54 2006 1347 661 49 2007 712 122 17 2008 262ACCEPTED5 2318 MANUSCRIPT88 2009 1934 1379 71 2010 1470 867 59 2011 6030 4310 71 2012 763 563 74 2013 185 180 97 2014 5296 608 11 Total 100646 57090 Meanb 2873 1629 57 1σb 4621 4317 a Denotes incomplete year or data gap b Excluding 1978 and 1995
120 ACCEPTED MANUSCRIPT
Table 4. Explosive volcanic SO2 emissions per Volcanic Explosivity Index (VEI) class a VEI Eruptions Detected Measured SO2 Est. undetected SO2 (%) Mass Flux Mean (Tg) 1σ (Tg) Mass (Tg) Flux (Tg) (Tg/yr)b (Tg/yr)b 0 222 0 - - - - 0.22 0.006 1 507 4 0.26 0.007 0.006 0.006 2.9 0.081 2 479 13 3.22 0.09 0.02 0.02 8.35 0.23 3 162 50 8.18 0.23 0.05 0.11 5.67 0.16 4 26 100 13.05 0.36 0.45 0.75 0 0 5 5 100 11.28 0.31 2.26 2.82 0 0 6 1 100 20 0.56 20 - 0 0 Total 1402 55.97 17 0.48 a. Extracted from the Global Volcanism Program Volcanoes of the World 4.3.0 database (accessed 19 Sept 2014) for Oct 1978 – Sep 2014. Listed eruptions without an assigned VEI (75 eruptions) were ignored. b. Mean annual SO2 flux for the ~36 year period of observations.
Table 5. Total volcanic SO2 emissions per Volcanic Explosivity Index (VEI) class a VEI Eruptions Detected Measured SO2 (%) Mass Flux Mean (Tg) 1σ (Tg) (Tg) (Tg/yr)b 0 222 5 6.6 0.18 0.6 1.28 1 507 4 4.6 0.13 0.05 0.15 2 479 13 21.4 0.59 0.07 0.3 3 162 50 22.4 0.62 0.14 0.45 4 26 100 13.2 0.37 0.44 0.74 5 5 100 11.3 0.31 2.26 2.82 6 1 100 20 0.56 20 - Total 1402 99.6 a. Extracted from the Global Volcanism Program Volcanoes of the World 4.3.0 database (accessed 19 Sept 2014) for Oct 1978 – Sep 2014. Listed eruptions without an assigned VEI (75 eruptions) were ignored. b. Mean annual SO2 flux for the ~36 year period of observations.
ACCEPTED MANUSCRIPT
121 ACCEPTED MANUSCRIPT
Table 6. Sources of eruptive volcanic SO2 emissions (>10 kt), 1978-2014 Total SO 2 Degassing Volcano emission Eruptions style* (Tg) Nyamuragira 25.149 17 1 Pinatubo 20.194 1 0 El Chichon 8.090 1 0 Sierra Negra 4.452 2 0 Holuhrauna 4.305 1 0 Hudson 4.001 2 0 Nabro 3.65 1 0 Anatahan 2.818 2 1 Kasatochi 2.000 1 0 Fernandina 1.821 5 0 Galunggungb 1.730 24 0 Cerro Azul 1.229 3 0 Sarychev Peak 1.200 1 0 Mauna Loa 1.197 1 0 Kilaueab 1.147 39 3 Alaid 1.110 2 0 Wolf 1.080 1 0 Mt St Helens 0.910 1 1 Hekla 0.883 4 0 Chikurachki 0.830 7 0 Spurr 0.800 3 0 Ruiz 0.762 3 3 Krafla 0.685 3 0 Rabaul 0.531 3 2 Etna 0.505 51 3 Copahue 0.500 1 1 Eyjafjallajökull 0.466 1 0 Soputan 0.455 19 1 Lascar 0.455 3 3 Soufriere Hillsb 0.450ACCEPTED 17 3 MANUSCRIPT Manam 0.420 12 2 Kliuchevskoi 0.403 6 1 Fogoa 0.382 1 0 Kelut 0.350 2 0 Grimsvötn 0.350 4 0 Ulawun 0.347 5 2 P. de la Fournaise 0.334 9 0 Pagan 0.320 1 1 Redoubt 0.315 2 1 Merapi 0.310 2 2 Banda Api 0.250 1 0 Zhupanovsky 0.227 1 0 Colo 0.200 1 0 Marchena 0.200 1 0
122 ACCEPTED MANUSCRIPT
Cordon Caulle 0.200 1 0 Tolbachik 0.200 1 0 Dalaffilla 0.150 1 0 Okmok 0.150 1 0 Lopevi 0.149 5 3 Pavlof 0.135 4 1 Ambrym 0.122 3 3 Asama 0.101 2 3 Sangeang Api 0.100 1 0 Tungurahua 0.090 5 3 Ruapehu 0.090 2 1 Popocatepetlb 0.087 5 3 Reventador 0.084 1 2 Ruang 0.080 1 0 Jebel at Tair 0.080 1 0 Shishaldin 0.063 1 1 Kizimen 0.060 2 0 Manda Hararo 0.060 2 0 Makian 0.050 1 0 Karkar 0.050 1 1 Mayon 0.050 1 3 Cameroon 0.048 2 0 Miyakejima 0.043 1 3 Karthala 0.041 4 0 Llaima 0.040 2 1 Paluweh 0.040 1 1 Bromo 0.031 1 3 Augustine 0.030 2 1 Nyiragongo 0.030 1 3 Bezymiannyb 0.026 5 1 Home Reef 0.025 1 0 Pacaya 0.025 2 3 Oshima 0.022 1 1 Lokon-Empung 0.020ACCEPTED 1 3 MANUSCRIPT Sinabung 0.020 1 1 H. T. Hungaapai 0.014 1 0 Gamalama 0.013 2 1 Sheveluchb 0.012 2 3 Santa Ana 0.012 1 3 * Qualitative assessment of degassing style: 0: Entirely eruptive degassing. No significant passive SO2 degassing known or reported. 1: Some passive degassing likely/measured/reported but eruptive degassing dominant 2. Approximately equal passive/eruptive degassing 3. Some eruptive degassing likely/measured/reported but passive degassing dominant. a. Preliminary estimate of SO2 emissions through the end of 2014. b. ‗Eruptions‘ listed correspond to discrete eruptive events producing the measured SO2 (although only a single long-lived eruption may be recorded in the VOTW 4.3 database).
123 ACCEPTED MANUSCRIPT
Table 7. HCl and SO2 mixing ratios measured by Aura/MLS in volcanic clouds
Volcano Eruption Date HCl (ppbv) SO2 (ppbv) Pressure (hPa) HCl/SO2 Manam Jan 27, 2005 4-6 279 68-100 0.01 Anatahan Apr 6, 2005 3.5 133 100 0.02 Soufriere Hills May 20, 2006 3 200 68 0.01 Chaitén May 6, 2008 1.6 28 147 0.03 Okmok Jul 12, 2008 5 212 147 0.01-0.02 Kasatochi Aug 7, 2008 5-6 392 68-215 0.01-0.014 Redoubt Mar 26, 2009 4-5 175 100-215 0.02 Sarychev Peak Jun 15, 2009 7-9 529 32-215 0.03 Merapi Nov 5, 2010 6-7 172 100-215 0.03 Cordón Caulle1 Jun 4, 2011 2-3 77 147-215 0.03 Nabro Jun 15, 2011 9 306 46-215 0.03 Paluweh Feb 2, 2013 2.2 129 100-215 0.01 Kelut Feb 14, 2014 7 398 32-147 0.01 Sangeang Api May 31, 2014 2-3 53 146 0.03 1. Data from Theys et al. (2014)
ACCEPTED MANUSCRIPT
124 ACCEPTED MANUSCRIPT
Figure captions
Figure 1. Satellite measurements of SO2 emissions by volcanic eruptions between October 1978 and April 2015. Measurements are primarily based on UV TOMS, OMI and OMPS data, supplemented by IR data from TOVS, AIRS and IASI. Eruptions are color-coded by estimated plume altitude, derived from a variety of sources, including Smithsonian Institution Global
Volcanism Program volcanic activity reports, volcanic ash advisories, and satellite data (e.g.,
CALIOP lidar for eruptions since 2006). Upper panel shows SO2 mass loading against time, with annual total explosive volcanic SO2 production (omitting SO2 discharge from effusive eruptions) shown in black. Orange lines above the plot indicate the operational lifetimes of the UV satellite instruments: Nimbus-7 (N7), Meteor-3 (M3), ADEOS (AD), and Earth Probe (EP) TOMS, OMI
(currently operational) and OMPS (currently operational). The TOMS data gap in 1995-96 is shown in gray (but note that TOVS SO2 data provide coverage during this time). Lower panel shows eruption latitude against time, with symbol size proportional to SO2 mass loading.
Horizontal dashed lines delineate the tropics. The charts only depict SO2 emissions from discrete volcanic eruptions; continuous emissions from passive degassing and some smaller eruptions are not included. Over 600ACCEPTED eruptions are shown, releasingMANUSCRIPT a total of ~100 Tg of SO2. The database shown in this plot will soon be available from the NASA Goddard Earth Sciences (GES) Data and Information Services Center (DISC) as a level 4 MEaSUREs (Making Earth System Data
Records for Use in Research Environments) data product.
Figure 2. Examples of A-Train synergy for characterizing volcanic cloud altitude. (a) OMI total column SO2 and MLS UTLS column SO2 (W. Read, pers. comm.) for the April 6, 2005 eruption
125 ACCEPTED MANUSCRIPT
of Anatahan (CNMI; triangle). MLS retrievals are shown as filled circles along the Aura satellite daytime and nighttime overpass tracks; note that circle size is not representative of the MLS field-of-view. Whilst the nadir OMI measurements detect SO2 east and west of Anatahan, MLS only detects UTLS SO2 to the east, confirming the variable altitude of the mapped SO2. (b)
Composite of nighttime A-Train measurements of the Chaitén volcanic plume on May 7, 2008 at
~4:47 UTC. Main panel shows 532 nm total attenuated backscatter (km-1 sr-1) from CALIOP, clearly showing aerosol in the layered volcanic cloud, overlain with reflectivity from the
CloudSat Cloud Profiling Radar (CPR), which in this case only detects hydrometeors in the tropospheric meteorological clouds (indicating that the volcanic aerosol particle size is below the radar detection limit, i.e., sub-mm). White stars show locations of SO2 detection by AIRS based on a 2-channel brightness temperature difference (BTD in K). The sub-horizontal white line indicates the tropopause altitude. MLS retrievals of SO2 (ppbv), HCl (ppbv) and ice-water content (IWC; mg m-3) for several pressure levels (indicated top left in hPa) are plotted above the
CALIOP data; these show clear enhancements in the volcanic cloud above 215 hPa.
Figure 3. SO2 mass loadings (kilotons; log scale) derived from the operational OMI Linear Fit
SO2 algorithm (YangACCEPTED et al., 2007) against those MANUSCRIPT derived from the Prata and Bernardo (2007)
AIRS SO2 algorithm for selected volcanic eruptions in the 2004-2014 period. Eruptions are color-coded by estimated plume altitude (see text for discussion of sources) and 20% error bars are shown. Dashed line is the 1:1 line. Note broad correlation between OMI and AIRS SO2 retrievals with greater scatter for smaller eruptions and/or lower altitude plumes, as expected.
126 ACCEPTED MANUSCRIPT
Figure 4. Time-series of maximum daily OMI SO2 index (OSI) and MLS UTLS SO2 and HCl measurements from September 2004 – December 2014. OMI panels (red curve) show the maximum daily value of the OSI recorded in global OMI measurements. The OSI is sensitive to
SO2 loading and altitude (see text for definition). Some of the prominent volcanic eruptions responsible for OSI excursions are labeled. MLS panels show the maximum retrieved UTLS SO2 and HCl mixing ratios on a pressure grid extending from 215 hPa (~11 km altitude) to 40 hPa
(~20-21 km altitude) for SO2 and from 150 hPa (~14 km altitude) to 40 hPa for HCl (note that although MLS HCl measurements are deemed invalid for scientific use below 100 hPa (Livesey et al., 2011), and clearly biased high at 147 hPa, it is clear that volcanic HCl is detected at this level). Note that noise in MLS measurements increases with pressure (i.e., towards lower altitudes) and the data shown here do not represent the full vertical extent of MLS measurements.
Blank regions are MLS data gaps. MLS data have been screened using the criteria recommended by Livesey et al. (2011).
Figure 5. Time-series of UTLS SO2 emissions (Tg, log scale) detected by AIRS in 2003-2014 using the Prata and Bernardo (2007) SO2 retrieval technique. Sources of the observed SO2 emissions are indicated.ACCEPTED SO2 loadings are also MANUSCRIPTcolored based on the color bar to the right of the plot.
Figure 6. Time-series (upper panel) and cumulative map (lower panel) of SO2 emissions detected by AIRS in 2011 (data for other years in 2003-2014 are provided as Supplementary
Material). Upper panel shows daily SO2 amounts (Tg) retrieved by AIRS using the Prata and
Bernardo (2007) algorithm (gray bars) and reported volcanic eruptions (colored bars and stars;
127 ACCEPTED MANUSCRIPT
color denotes VEI). Horizontal bars indicate eruption duration reported by the Smithsonian
GVP; horizontal arrows indicate eruptions or activity continuing into the following year. Note that not all the indicated eruptions produced SO2 detected by AIRS. Lower panel shows cumulative SO2 column amounts retrieved by AIRS with major eruptions indicated.
Figure 7. Time-series of IASI total SO2 mass (lower panel; logarithmic scale in kilotons) and altitude-binned SO2 mass (upper panel; for altitudes >5 km only) measurements from October
2007 – December 2014, derived using the retrieval techniques described by Clarisse et al. (2012,
2014). Prominent volcanic eruptions are labeled in the lower panel.
Figure 8. Average SO2 columns measured by IASI over ~7 years (October 2007 – December
2014). The scale is logarithmic (log[1+(mean SO2 column in DU)]). The more prominent eruptions evident in this map include Kasatochi (Alaska; August 2008), Sarychev Peak (Kurile
Islands; June 2009), Merapi (Indonesia; November 2010), Grimsvötn (Iceland; May 2011),
Cordón Caulle (Chile; June 2011), Nabro (Eritrea; June 2011) and Copahue (Chile; December
2012).
ACCEPTED MANUSCRIPT
Figure 9. Cumulative SO2 emissions from volcanic eruptions against time for 1978-2015, showing emissions for all eruptions detected from space (blue line), explosive eruptions only
(red line) and effusive eruptions only (green line). Black line shows global stratospheric aerosol optical depth (SAOD) at 550 nm for 1978-2012 from Sato et al. (1993) and recent updates through December 2012 (http://data.giss.nasa.gov/modelforce/strataer/); gray line shows the tropical (30ºS-30ºN) SAOD at 525 nm data of Vernier et al. (2011; with recent updates) for
128 ACCEPTED MANUSCRIPT
1985-2014 based on SAGE II, GOMOS and CALIPSO observations of aerosol between 15 and
40 km altitude. Differences in the SAOD datasets arise from different sources of SAOD observations, interpolation methods, latitude range (the Sato et al. (1993) data plotted are global whereas the Vernier et al. (2011) data are for the tropics; thus high-latitude eruptions produce a weaker signal in the latter) and altitude ranges (the Vernier et al. (2011) data cover altitudes of
15-40 km and hence includes part of the tropical upper troposphere, resulting in larger AODs in some periods).
Figure 10. Daily maximum UTLS mixing ratios of SO2 (upper panel) and HCl (lower panel) measured by MLS between June 10 and July 30, 2011, showing the impact of the June 12, 2011 eruption of Nabro (Eritrea).
Figure 11. Upper panel: SO2 mass loading against Volcanic Explosivity Index (VEI) for all measured eruptions (1978-2014). Triangles and circles denote eruptions detected by TOMS (i.e., in 1978-2004) and OMI (i.e., in 2004-2014), respectively. Lower panel: SO2 mass loading against VEI for explosive eruptions only (i.e., excluding effusive eruptions regardless of VEI).
VEIs are derived fromACCEPTED the Smithsonian Global MANUSCRIPT Volcanism Program (GVP) database (VOTW
4.3.0).
Figure 12. Global map of sources of eruptive SO2 emissions detected from space in 1978-2014.
Symbol size is proportional to total SO2 emissions measured for each volcano (Table 6), which is also indicated by the symbol color scale. Triangles indicate explosive eruptions and circles indicate effusive eruptions.
129 ACCEPTED MANUSCRIPT
Figure 13. Time-series of SO2 mass loading similar to Figure 1, but with only eruptions reaching altitudes of ≥15 km shown. The time axis is extended to include the 1974-1978 period prior to the satellite remote sensing era. Orange line shows stratospheric aerosol optical depth (SAOD) at
550 nm from Sato et al. (1993) and recent updates through December 2012
(http://data.giss.nasa.gov/modelforce/strataer/); gray line shows the tropical (30ºS-30ºN) SAOD at 525 nm data of Vernier et al. (2011; with recent updates) for 1985-2014 based on SAGE II,
GOMOS and CALIPSO observations of aerosol between 15 and 40 km altitude. Differences between the SAOD datasets are explained in Figure 9.
Figure 14. Altitude of SO2 injection against e-folding time for SO2 removal for selected volcanic eruptions and emissions since 1980, with e-folding times derived from the literature (pre-2007) and from IASI SO2 measurements (2007-2014). Symbol size is proportional to total SO2 emission, and symbol color indicates volcano latitude. The SO2 emissions represented are from:
Erebus 1980 (Radke et al., 1982), Nyamuragira 1981 (Krueger et al., 1996; Bluth and Carn,
2008), Nyamuragira 1980-2004 (Bluth and Carn, 2008), El Chichón 1982 (Bluth et al., 1997),
Pinatubo 1991 (BluthACCEPTED et al., 1997), Etna 1983 (MANUSCRIPTMartin et al., 1986), Redoubt 1990 (Hobbs et al.,
1991), Cerro Hudson 1991 (Constantine et al., 2000), Soufrière Hills 1996 (Oppenheimer et al.,
1998), Masaya 2003 (McGonigle et al., 2004), Soufrière Hills 2004 (Rodriguez et al., 2008),
Jebel at Tair 2007 (Clarisse et al., 2008), Kasatochi 2008 (Krotkov et al., 2010), Sarychev Peak
2009 (Haywood et al., 2010), Cordon Caulle 2011 (Theys et al., 2013), Merapi 2010 (IASI derived), Grimsvötn 2011 (IASI derived), and Copahue 2012 (IASI derived). Several e-folding
130 ACCEPTED MANUSCRIPT
times below 0.1 days were reported by Oppenheimer et al. (1998) for lower tropospheric SO2 plumes but are not shown here.
Figure 15. Sources of passive volcanic SO2 emissions detected by OMI: (a) Most persistent sources (in rank order) detected in OMI data collected between September 2004 and December
2013; (b) Most persistent sources detected by OMI in 2014. Colors correspond to the average value of the OSI for each volcano. Note that for some volcanoes in close proximity, the source of observed SO2 emissions can be ambiguous (e.g., Nyiragongo-Nyamulagira, DR Congo;
Mutnovsky-Gorely, Kamchatka).
Figure 16. Latitude and strength of persistent volcanic SO2 sources against time based on OMI
SO2 measurements from September 2004 – December 2014. For each day, the latitudes of all
OMI pixels that show the highest daily values of the OMI SO2 Index (OSI; see section 3.4 for definition) and are located within 50 km of an active volcano are plotted. Persistent SO2 sources thus appear as horizontal bands on this plot. Although most eruptive SO2 emissions are excluded from this plot (see section 3.4), short discontinuities in the data are due to large SO2-rich eruptions temporarilyACCEPTED overwhelming the OSI measurementsMANUSCRIPT (e.g., the Kasatochi eruption in
August 2008, the Sarychev Peak eruption in June 2009, and the Nabro eruption in June 2011).
131 ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Fig. 1
132 ACCEPTED MANUSCRIPT
Fig. 2a
ACCEPTED MANUSCRIPT
133 ACCEPTED MANUSCRIPT
Fig. 2b
ACCEPTED MANUSCRIPT
134 ACCEPTED MANUSCRIPT
Fig. 3
ACCEPTED MANUSCRIPT
Fig. 4
135 ACCEPTED MANUSCRIPT
Fig. 5 ACCEPTED MANUSCRIPT
136 ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Fig. 6
137 ACCEPTED MANUSCRIPT
Fig. 7
ACCEPTED MANUSCRIPT
138 ACCEPTED MANUSCRIPT
Fig. 8 ACCEPTED MANUSCRIPT
139 ACCEPTED MANUSCRIPT
Fig. 9
ACCEPTED MANUSCRIPT
Fig. 10
140 ACCEPTED MANUSCRIPT
Fig. 11
ACCEPTED MANUSCRIPT
141 ACCEPTED MANUSCRIPT
Fig 12
ACCEPTED MANUSCRIPT
Fig. 13
142 ACCEPTED MANUSCRIPT
Fig. 14
ACCEPTED MANUSCRIPT
143 ACCEPTED MANUSCRIPT
Fig. 15a
Fig. 15b
ACCEPTED MANUSCRIPT
Fig. 16
144